Feature agents-write-git, step 2: git over SSH with a key made on each machine (issue 955) #340

Merged

mesh-tools merged 14 commits from feat/agents-write-git-ssh into main 2026-10-16 22:31:17 +00:00

Conversation 4 Commits 14 Files Changed 26

+2687 -121

mesh-tools commented 2026-10-16 20:58:34 +00:00

Member

Copy Link

Copy Source

Issue 955, under ADR 0399 and ADR 0400 (accepted 2026-10-16).

ssh-client (every machine): its tools bundle keeps ~/.ssh/mesh_ed25519 for the machine's account, made there when missing and never overwritten, and writes the public half and fingerprint to its own state keys at <node>.<account> (no private material). It requires forge-account, and its templates fill a marked block in known_hosts, ~/.ssh/config.d/10-forge, and a marked block in ~/.config/git/config that rewrites https://git.novox.be/ to ssh://git@novox.internal:222/.

gitea: a new forge-account-provider (SDK provider.New, with gitea's own idempotent Make and Gone) reads ssh-client.keys through a reads grant, keeps agent-<node> in the agents' team, registers exactly the reported key under the title mesh:<node>.<account> (removing earlier mesh: keys, never another key), and serves user, ssh_host, ssh_port, known_host, ssh_base and https_base in its bound entry. It removes nothing until the key state has been read once. The agent-account code moves to internal/forgeaccount, shared with agent-token.

Not in this step: retiring the claude-code token (ADR 0399 §4, step 3).

Review: 3 rounds by the reviewer on the larger model, approved at 6f45570d4c. Integration tests S4 (key process over a real bus) and F5 (provider over a real bus, fake forge) run in the merge check.

Opened along the way: 965, 966, 967.

One delivery group with novox/hq feat/agents-write-git-ssh (design 29).

Issue 955, under ADR 0399 and ADR 0400 (accepted 2026-10-16). **ssh-client** (every machine): its tools bundle keeps `~/.ssh/mesh_ed25519` for the machine's account, made there when missing and never overwritten, and writes the public half and fingerprint to its own state `keys` at `<node>.<account>` (no private material). It requires `forge-account`, and its templates fill a marked block in `known_hosts`, `~/.ssh/config.d/10-forge`, and a marked block in `~/.config/git/config` that rewrites `https://git.novox.be/` to `ssh://git@novox.internal:222/`. **gitea**: a new `forge-account-provider` (SDK `provider.New`, with gitea's own idempotent Make and Gone) reads `ssh-client.keys` through a `reads` grant, keeps `agent-<node>` in the agents' team, registers exactly the reported key under the title `mesh:<node>.<account>` (removing earlier `mesh:` keys, never another key), and serves `user`, `ssh_host`, `ssh_port`, `known_host`, `ssh_base` and `https_base` in its `bound` entry. It removes nothing until the key state has been read once. The agent-account code moves to `internal/forgeaccount`, shared with agent-token. Not in this step: retiring the claude-code token (ADR 0399 §4, step 3). Review: 3 rounds by the reviewer on the larger model, approved at 6f45570d4c23f5248144e2ffef5d4c7e920ba086. Integration tests S4 (key process over a real bus) and F5 (provider over a real bus, fake forge) run in the merge check. Opened along the way: 965, 966, 967. One delivery group with novox/hq feat/agents-write-git-ssh (design 29).

mesh-tools added 13 commits 2026-10-16 20:58:34 +00:00

Lift the agent account code into internal/forgeaccount for the forge-account provider (issue 955 F1) c8dfbdc813

Register and prune an agent account's mesh SSH keys through the forge API (issue 955 F2) 8131c13634

Keep the machine's forge SSH key and its keys entry (issue 955 S1) b2009f1bb5

Declare ssh-client's keys state and its forge-account requirement (issue 955 S2) a4cdd05cb2

Place the forge's known host, SSH Host and git rewrite (issue 955 S3) ad5a0f8e44

Prove the forge key process over a real bus in the merge check (issue 955 S4) 9cb615f121

Merge branch 'feat/agents-write-git-ssh' into feat/agents-write-git-ssh-s 165cbcb35f

Add the forge-account provider: an agent account, its team and its machine key per consumer (issue 955 F3) c346bba34e

Declare the forge-account provision, its provider bundle, its fact and the keys it reads (issue 955 F4) 1081a5f3ea

Test the forge-account provider whole against a real bus and a fake forge (issue 955 F5) 5c64b71cd7

Merge feat/agents-write-git-ssh into feat/agents-write-git-ssh-f (issue 955 F) 896f1d2854

Remove a machine's mesh keys and entry when its key is no longer reported (issue 955 F) 6fdf3629d3

Do not remove keys or entries before the machines' keys are read (issue 955 F) 6f45570d4c

mesh-tools commented 2026-10-16 20:58:48 +00:00

Author

Member

Copy Link

Copy Source

Delivery novox/mesh-catalog@ae9cda8531b3 — failed since 2026-10-16T22:32:35Z

Covered: covered after its failure: a later walk that contains it was delivered: plan-1792195559895153198 (novox/mesh-catalog at 82df3bd5 into main, done, contains its merge 734a77f5)

Delivery plan — builds gitea, ssh-client → ace, g14, novox, shanks; no bus step

• tier 0: gitea, ssh-client
• ace: receives ssh-client
• g14: receives ssh-client
• novox: receives gitea, ssh-client
• shanks: receives ssh-client
• gitea provides npm-package-registry, git, gitea-read: its consumers are sent again after it
• gitea keeps data (ADR 0233): what it holds is backed up before it moves
• ssh-client keeps data (ADR 0233): what it holds is backed up before it moves
• grants step: if this changes what the bus's user list says (a new tool, subject or user), novox is sent that list alone, every build there kept, before ssh-client is judged on its first machine

Group feat/agents-write-git-ssh, in order: novox/hq@24701df23e → novox/mesh-catalog@ae9cda8531

• composed together: pass — every machine composes with the change as it did without (4 of 4 compose)

Transitions

• 2026-10-16 22:05 (new) → proposed (announced): novox/mesh-catalog#340's head announced
• 2026-10-16 22:08 proposed → checked (checked): a verdict names this commit and the trunk's head as its trunk commit
• 2026-10-16 22:08 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail
• 2026-10-16 22:31 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move
• 2026-10-16 22:31 published → delivering (go): its walk started: let go by this owner in its group's order, by a person, or on the controller's own path — not a failed walk that carried a later commit for it, after which it waits for a walk of its own commit
• 2026-10-16 22:32 delivering → failed (failed): its walk failed, carrying its own commit: a gate on a first machine (what it carried put back), a build, a machine

The commit's note under refs/notes/mesh-plan keeps every transition: git log --notes=mesh-plan.

<!-- mesh-delivery:view --> **Delivery** `novox/mesh-catalog@ae9cda8531b3` — **failed** since 2026-10-16T22:32:35Z **Covered**: covered after its failure: a later walk that contains it was delivered: plan-1792195559895153198 (novox/mesh-catalog at 82df3bd5 into main, done, contains its merge 734a77f5) **Delivery plan** — builds gitea, ssh-client → ace, g14, novox, shanks; no bus step - tier 0: gitea, ssh-client - ace: receives ssh-client - g14: receives ssh-client - novox: receives gitea, ssh-client - shanks: receives ssh-client - gitea provides npm-package-registry, git, gitea-read: its consumers are sent again after it - gitea keeps data (ADR 0233): what it holds is backed up before it moves - ssh-client keeps data (ADR 0233): what it holds is backed up before it moves - grants step: if this changes what the bus's user list says (a new tool, subject or user), novox is sent that list alone, every build there kept, before ssh-client is judged on its first machine **Group** `feat/agents-write-git-ssh`, in order: novox/hq@24701df23e8a → novox/mesh-catalog@ae9cda8531b3 - composed together: pass — every machine composes with the change as it did without (4 of 4 compose) **Transitions** - 2026-10-16 22:05 (new) → proposed (announced): novox/mesh-catalog#340's head announced - 2026-10-16 22:08 proposed → checked (checked): a verdict names this commit and the trunk's head as its trunk commit - 2026-10-16 22:08 checked → ready (accepted): the gate passed or warned, and the repository's own check did not fail - 2026-10-16 22:31 ready → published (merged): on the trunk its modules follow: merged there, and its walk opened — or nothing for a walk to move - 2026-10-16 22:31 published → delivering (go): its walk started: let go by this owner in its group's order, by a person, or on the controller's own path — not a failed walk that carried a later commit for it, after which it waits for a walk of its own commit - 2026-10-16 22:32 delivering → failed (failed): its walk failed, carrying its own commit: a gate on a first machine (what it carried put back), a build, a machine The commit's note under `refs/notes/mesh-plan` keeps every transition: `git log --notes=mesh-plan`.

mesh-tools commented 2026-10-16 21:02:40 +00:00

Author

Member

Copy Link

Copy Source

Merge check at 6f45570d
Judged: head 6f45570d4c23 merged into trunk 017b729bc95d; read beside it: mesh-controller 53a3c91f53ce, mesh-controller-main 1cdf12df00b5, mesh-host cd13f7726ef0, mesh-tools 7412c5b6c9c5

• mesh/merge-gate: FAIL — 4 problem(s) the change brings; the first: a machine (f35): nothing could be sent to it with this change — ssh-client asks its forge-account binding for "known_host", and what answers it says "address", "as", "at" or "from" (modules: gitea, ssh-client)
• mesh/repo-check: PASS — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)

Delivery plan — builds gitea, ssh-client → ace, g14, novox, shanks; no bus step

• tier 0: gitea, ssh-client
• ace: receives ssh-client
• g14: receives ssh-client
• novox: receives gitea, ssh-client
• shanks: receives ssh-client
• gitea provides npm-package-registry, git, gitea-read: its consumers are sent again after it
• gitea keeps data (ADR 0233): what it holds is backed up before it moves
• ssh-client keeps data (ADR 0233): what it holds is backed up before it moves
• grants step: if this changes what the bus's user list says (a new tool, subject or user), novox is sent that list alone, every build there kept, before ssh-client is judged on its first machine
• read by no module's build: modules/ssh-client/cmd/ssh-client-tools/keykeeper_bus_test.go, modules/ssh-client/cmd/ssh-client-tools/keykeeper_test.go, modules/ssh-client/cmd/ssh-client-tools/templates_test.go

Run by the build seat on shanks as build-1792184317773818513 (builds --log build-1792184317773818513).

✔ the accounts that may mark are the setting's, and only the mesh's own when it cannot be read (0.996036ms)
✔ the newest status of each context is kept, with what it said, when and by whom (29.49776ms)
✔ a commit merges only when every required status is a success; a warning blocks (1.750482ms)
✔ no rule requires nothing; a rule that cannot be read says nothing, never yes (0.327742ms)
✔ a commit is contained in another when the compare finds nothing on it the other lacks (0.983136ms)
✔ a note line is appended once, as the forge's user, in its own repository (21.307318ms)
✔ a note that cannot be written is said, never read as written (0.537074ms)
✔ one view per pull request, found by its marker; only the mesh's statuses (0.779275ms)
✔ a build machine taking the check is said on the head: building on which machine, since the local time (37.301421ms)
✔ the verdict heard while a 'building' write is in flight lands last and stays (6.700263ms)
[gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check now pending; not written
✔ the verdict written while a start's read of the forge is in flight stays the newest status (1.43382ms)
✔ the verdict heard while a 'building' write still waits its turn: the 'building' is never written (0.628414ms)
✔ a start heard after the verdict writes nothing (0.611684ms)
✔ a start from before the check now pending (a run before a recheck, the stream's history) writes nothing (0.804865ms)
✔ an old run's verdict redelivered after a recheck's pending is not written over it (1.128977ms)
[gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check the forge holds pending; not written
[gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check now pending; not written
✔ an old run's verdict in flight when the recheck comes lands before the recheck's pending (3.056779ms)
✔ a rechecked head says where its new run is: building on which machine (1.184608ms)
✔ the recheck's own verdict is written, even asked a little before the pending by another machine's clock (0.832626ms)
✔ after a restart, an old run's verdict is weighed against the check the forge's history holds open (0.665614ms)
✔ a recheck the forge does not take fails its caller, and the head is as it was (0.598733ms)
✔ the status tool writes the merge check's pending through the writer, and the mesh's other statuses as they were (5.590877ms)
✔ when a check was asked is read from its id; the opening of the check now pending from the gate's history (0.408703ms)
✔ a head this process did not announce (it restarted) is said only while the forge holds its gate pending (1.52964ms)
✔ a repository as the mesh clones it is told from another (0.361912ms)
✔ gitea_api refuses the forge's administration, a branch's protection and the merge check's status, and sends nothing (6.1449ms)
✔ gitea_api refuses a file written to the default branch or one a rule guards, and writes one on a branch of its own (55.81583ms)
✔ a protection rule covers a branch as the forge decides it: classes, alternatives, a plain name in any case (issue 539) (30.288405ms)
✔ gitea_api still writes what an author writes (9.915634ms)
✔ gitea_branch_protection_set only strengthens (20.1742ms)
✔ gitea_merge_pull_request merges only a head whose required statuses all succeeded, and only that head (50.631746ms)
✔ gitea_merge_pull_request refuses a verdict judged against an earlier trunk, and asks the check once (150.49121ms)
✔ gitea_merge_pull_request merges only the head it is named, and refuses one that is not the head (17.699374ms)
✔ gitea_merge_pull_request is loud when the forge merged a head other than the one it was handed (13.628448ms)
✔ gitea_merge_pull_request says the merge happened when reading back what landed fails (25.294584ms)
✔ gitea_merge_pull_request reads back a fast-forward, and says it cannot read back a squash or a rebase (57.836283ms)
✔ gitea_delete_repo deletes only the tools' own user's repository, unguarded, untouched for 30 days, named in full (21.050976ms)
✔ gitea_api sends the path it judged, and refuses moving a pull request's base (3.622374ms)
✔ gitea_merge_pull_request merges only into a branch whose protection requires the checks (7.291017ms)
✔ a rule named for a branch is not made weaker than the glob rule that guards it now (13.722928ms)
✔ a repository owner or name is never another path (0.309732ms)
✔ the judges, without a forge (0.381263ms)
✔ main's protection is never weakened in its push or its merge whitelist (issue 948, ADR 0398 §2) (0.248002ms)
✔ gitea_api never acts as another user: sudo, in any case, and any query key it does not send are refused (1.783772ms)
✔ gitea_api never approves a pull request nor requests changes: those are the operator's (2.503066ms)
✔ no tool's caller text becomes another path or another user: `..` and `?sudo=` are refused before the forge (5.113573ms)
✔ gitea_merge_pull_request merges a verdict that read siblings only with its group passed and each sibling head reviewed (89.879421ms)
✔ a verdict put out of date by a sibling's move is refused in its own words, and asked again once (23.793683ms)
✔ every page of a pull request's files is read, though the forge caps a page at fifty (39.192863ms)
✔ a pass asks only the repositories that moved since the last look, every one before the first (0.614393ms)
✔ the directories above a merge's files that hold a module at the commit are said, the root never (18.087636ms)
✔ a lookup the forge refuses is thrown, not read as no module (5.740277ms)
✔ a renamed file is said under its old name too, as gone (novox/hq ADR 0267) (9.399361ms)
✔ a new head beyond the newest 20 open pull requests is announced, exactly once (75.930099ms)
✔ a pass pages until it reaches back past its floor, and no further (31.20745ms)
✔ gitea_branch_protection_get answers every rule, read with the administrator account the tools' token cannot read with (29.149917ms)
✔ gitea_branch_protection_get answers the rule for a branch, null when it has none (10.188606ms)
✔ gitea_branch_protection_set still gets the forge's refusal, as the tools' account (4.105636ms)
✔ the statuses asked for replace the rule's, and an empty list requires none (5.421895ms)
✔ a new rule refuses direct pushes unless asked; an administrator's override only when said (0.394442ms)
✔ main's body always refuses pushes and lets only the tools' account merge, whatever is asked (0.858655ms)
✔ another branch's body carries no merge whitelist (0.287392ms)
✔ the rule main is put right in the forge's own words: the body a setup sends carries no status checks of its own (0.394792ms)
✔ each open pull request's new head is announced once, and nothing closed or merged (20.358261ms)
✔ a verdict is the head commit's status; an error is the forge's error, never a success (0.955257ms)
✔ each layer is its own status: the gate with the modules it judged, the repository's own check beside it (1.013266ms)
✔ no verdict sets warning on a required check; only a person's decision fails (0.922646ms)
✔ a commit status is set on the commit, under the merge check's context (43.049537ms)
✔ a change plan is the gate's result: said on the status and, when it builds something, as a comment (0.571763ms)
✔ a head announced is pending in the build queue, since the local time it was announced (0.350792ms)
✔ the vault's token is used, a setting of it is ignored and said once, without its value (2.573667ms)
[gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 at bbbbbbbbbbbb, it is now at dddddddddddd: out of date, nothing run (novox/hq ADR 0382)
[gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was closed unmerged: out of date, nothing run (novox/hq ADR 0382)
[gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 merged onto cccccccccccc, its main is now eeeeeeeeeeee: out of date, nothing run (novox/hq ADR 0382)
[gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was merged otherwise than its head bbbbbbbbbbbb onto cccccccccccc: out of date, nothing run (novox/hq ADR 0382)
[gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was merged otherwise than its head bbbbbbbbbbbb onto cccccccccccc: out of date, nothing run (novox/hq ADR 0382)
[gitea] novox/mesh-host's main moved from 444444444444 to 333333333333: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323)
✔ both statuses name the siblings the verdict read, keeping the verdict word and the start of its reason within 140 characters (2.593177ms)
✔ a verdict that could not learn its siblings says so on the status and in the comment, and names none (0.428613ms)
✔ a verdict without siblings is said exactly as before (an older controller) (0.856436ms)
✔ the siblings a verdict names are kept per member, replaced by its next verdict, and removed by one that read none (1.009537ms)
✔ a sibling's new head puts both statuses of the member back to pending, with the reason, and runs nothing (3.916835ms)
✔ a sibling closed unmerged puts the member's verdict out of date (0.503023ms)
✔ a move of the sibling's base puts it out of date, naming both commits (0.508053ms)
✔ the sibling's own merge of the head read onto the base read does not, in either order of the two merges (0.690205ms)
✔ a merge of another head, or onto another base, is a move (1.148408ms)
✔ only the members whose verdict names the sibling are touched, and a member that moved on is dropped (0.850135ms)
✔ a trunk that moved tells the siblings' weighing; one first seen does not (0.859435ms)
✔ the group rule names every missing thing: the group status at each head read, and each sibling head's review (1.073487ms)
✔ a merged or closed member's entry is dropped, so a sibling's later push never re-pends its merged head (0.270672ms)
✔ a consumer that asks nothing, or write, publishes: the package team (2.224155ms)
✔ a consumer that asks to read joins the readers, who write no package (0.720095ms)
✔ a reader is made, put on the readers team, and taken off the package team (0.549103ms)
✔ a publisher is made and put on the package team, as before (0.224721ms)
✔ first start: mints with the admin account, keeps the token at 0600, asks for the tools' scopes only (44.501937ms)
✔ second start: reuses the kept token, mints nothing (28.982286ms)
✔ the forge rejects the kept token (its data was restored): minted afresh, once, and the call goes through (28.976037ms)
✔ a token kept with write:admin by an earlier build is minted again without it at the next start, and the forge drops the old one (32.020797ms)
✔ a kept token lacking a scope the tools need: the forge refuses the call for the scope, the token is re-minted with the whole list, and the call goes through (27.696079ms)
✔ the kept file is gone but the forge still holds a token by that name: replaced, not refused (25.309174ms)
✔ concurrent first calls share one mint (16.507176ms)
✔ two tokens hold the name (two mints raced): both are deleted by their id, and one fresh token is minted (16.284155ms)
✔ two processes on one kept file, both without a token, mint once: the second waits for the first and reuses it (62.02393ms)
✔ a lock left by a process that died mid-mint is taken over, not waited on for ever (6.028509ms)
✔ the admin account is refused: said plainly, nothing kept, and the next call fails the same way rather than crashing (17.449353ms)
✔ one process shares one source per kept file — the watcher and the tools never renew against each other (0.814335ms)
✔ a second process finds the token the first renewed, and reuses it instead of minting over it (22.728606ms)
✔ a configured token wins, and is reported rather than minted over when the forge rejects it (2.704307ms)
✔ nothing to mint with and no token: the client says what is missing (0.213322ms)
[gitea] no tools — no Gitea token, and nothing to mint one with — set MESH_GITEA_TOKEN, or MESH_GITEA_ADMIN_USER, MESH_GITEA_ADMIN_PASSWORD_FILE, MESH_GITEA_STATE_DIR
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-dT9Z6L/state/token
✔ the tools register once there is a way to a token, and the first call mints it (29.48841ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-pR2QPr/state/token
✔ a pull request can be closed, read and cleaned up through the tools (37.673513ms)
[gitea] no token kept — minting one
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-V03qOP/state/token
✔ the tools' token is minted for an account of their own, which is no administrator and writes through its team (93.23407ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-6UD3o4/state/token
✔ with the tools' token, the forge writes a status and refuses a protection edit — the forge itself, not a guard (62.276681ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-2Ym0YR/state/token
✔ what a base's protection asks is read from the branch, which the tools' account may read, not from the rules' list it may not (50.752207ms)
[gitea] the kept token was minted with other scopes than the tools take now — minting one with them
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-aXncVG/state/token
[gitea] deleted 1 token(s) named "mesh-tools" the admin account "mesh-admin" held: the tools act as "mesh-tools" now
✔ the administrator's token from before goes: a kept admin token is minted again for the tools' account, and the forge drops the old one (52.544168ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-S7FEmb/state/token
✔ a tools' account someone made an administrator is made a plain account again at the next mint (38.297558ms)
✔ the tools never act as the administrator: MESH_GITEA_TOOLS_USER naming it is refused (1.012227ms)
✔ the manifest hands the tools their account, and the merge-ready authors name it rather than the administrator (1.042757ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-RH9eMW/state/token
✔ a repository check said as a warning on a head without merge-check.sh reads the base's protection from the branch (41.591149ms)
[gitea] no token kept — minting one
✔ a mint that fails leaves the administrator's old token in place: the tools are never left without one (31.326852ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-u4A0Fu/state/token
[gitea] FAILED to delete the token(s) named "mesh-tools" the admin account "mesh-admin" holds — tried again at the next start: Gitea DELETE /users/mesh-admin/tokens/902 (named mesh-tools): 500 {"message":"the forge is not well"}
✔ an old token that cannot be deleted after the mint is said, the new token kept, and deleted at the next start (49.043327ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-eVEpdU/state/token
✔ a team another process made between the look and the create is found, not a failure (42.530485ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-z9RksA/state/token
✔ setting the tools' account up puts main's rule right in every repository: no push, only the tools' account merges (issue 948) (38.526369ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-Dnd9F7/state/token
✔ a glob rule covering main keeps its statuses when main is given a rule of its own name (issue 948, review round 1) (38.187117ms)
✔ a start with the token kept still puts main's rules right: the module's start asks for it each time (issue 948, review round 1) (13.581238ms)
[gitea] novox/mesh-host's main moved from a2c2c35bbbbb to d28ca95ccccc: 25 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323)
[gitea] novox/mesh-host's main moved from d28ca95ccccc to ffffffffffff: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323)
[gitea] novox/mesh-host: whether a trunk moved could not be weighed, so it is weighed again on the next pass — the forge is away
[gitea] novox/mesh-host's main moved from a2c2c35bbbbb to d28ca95ccccc: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323)
[gitea] novox/mesh-controller's feat/57-config-and-provisioning moved from eeeeeeeeeeee to d28ca95ccccc: 1 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323)
✔ both statuses say the head and the trunk commit the verdict judged, first, whatever the summary's length (2.495656ms)
✔ a trunk's move puts an open head's merge check out of date, naming both commits, unless a check is on its way (1.4566ms)
✔ a push to a trunk puts every open pull request into it out of date — twenty-five of them — and none into another (5.841468ms)
✔ a trunk that cannot be read is weighed again on the next pass, never recorded half (0.951097ms)
✔ the trunk's move takes its turn after a verdict in flight, and weighs what the forge holds then (0.585294ms)
✔ a verdict judged against a trunk newer than the pass's first read is not put out of date (0.345292ms)
✔ a verdict that landed out of date keeps its account on the pull request (0.446332ms)
✔ a release branch's move puts every open pull request into it out of date, as a trunk's does (ADR 0358 §5) (1.464999ms)
✔ a signed event is published on the bus and acted on once (67.290554ms)
✔ an unsigned event, or one whose signature does not verify, is refused and published nowhere (39.054402ms)
✔ an event and the poll seeing the same head announce it once (0.563423ms)
✔ each forge event is named in the module's stream and trimmed (0.735795ms)
✔ the webhook is made a system webhook, which the forge sends every repository's events to (1.47756ms)
✔ default webhooks left at the receiver by an earlier start are replaced by one system webhook (1.428569ms)
✔ a repository's copy of a default webhook at the receiver is deleted, on every page, and said (5.287073ms)
✔ a forge that makes no system webhook is said, never taken as heard (0.993257ms)
✔ the receiver tells what it verified and what it refused (10.845129ms)
✔ a forge that never delivers is said loudly once, and unhealthy, when the poll finds what no event said (1.365349ms)
✔ a forge whose events stop after one delivery is said loudly once per silent spell, and refusals are unhealthy (0.532724ms)
✔ without a receiver by configuration the polls alone watch, healthy; a receiver that failed is unhealthy (0.434543ms)
✔ the polls look every 30 s until events are heard, every five minutes while they are, and 30 s again on a miss (0.258161ms)
ℹ tests 147
ℹ suites 0
ℹ pass 147
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 849.901758

**Merge check** at `6f45570d` Judged: head `6f45570d4c23` merged into trunk `017b729bc95d`; read beside it: mesh-controller `53a3c91f53ce`, mesh-controller-main `1cdf12df00b5`, mesh-host `cd13f7726ef0`, mesh-tools `7412c5b6c9c5` - `mesh/merge-gate`: **FAIL** — 4 problem(s) the change brings; the first: a machine (f35): nothing could be sent to it with this change — ssh-client asks its forge-account binding for "known_host", and what answers it says "address", "as", "at" or "from" (modules: gitea, ssh-client) - `mesh/repo-check`: **PASS** — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript) **Delivery plan** — builds gitea, ssh-client → ace, g14, novox, shanks; no bus step - tier 0: gitea, ssh-client - ace: receives ssh-client - g14: receives ssh-client - novox: receives gitea, ssh-client - shanks: receives ssh-client - gitea provides npm-package-registry, git, gitea-read: its consumers are sent again after it - gitea keeps data (ADR 0233): what it holds is backed up before it moves - ssh-client keeps data (ADR 0233): what it holds is backed up before it moves - grants step: if this changes what the bus's user list says (a new tool, subject or user), novox is sent that list alone, every build there kept, before ssh-client is judged on its first machine - read by no module's build: modules/ssh-client/cmd/ssh-client-tools/keykeeper_bus_test.go, modules/ssh-client/cmd/ssh-client-tools/keykeeper_test.go, modules/ssh-client/cmd/ssh-client-tools/templates_test.go Run by the build seat on shanks as `build-1792184317773818513` (`builds --log build-1792184317773818513`). ``` ✔ the accounts that may mark are the setting's, and only the mesh's own when it cannot be read (0.996036ms) ✔ the newest status of each context is kept, with what it said, when and by whom (29.49776ms) ✔ a commit merges only when every required status is a success; a warning blocks (1.750482ms) ✔ no rule requires nothing; a rule that cannot be read says nothing, never yes (0.327742ms) ✔ a commit is contained in another when the compare finds nothing on it the other lacks (0.983136ms) ✔ a note line is appended once, as the forge's user, in its own repository (21.307318ms) ✔ a note that cannot be written is said, never read as written (0.537074ms) ✔ one view per pull request, found by its marker; only the mesh's statuses (0.779275ms) ✔ a build machine taking the check is said on the head: building on which machine, since the local time (37.301421ms) ✔ the verdict heard while a 'building' write is in flight lands last and stays (6.700263ms) [gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check now pending; not written ✔ the verdict written while a start's read of the forge is in flight stays the newest status (1.43382ms) ✔ the verdict heard while a 'building' write still waits its turn: the 'building' is never written (0.628414ms) ✔ a start heard after the verdict writes nothing (0.611684ms) ✔ a start from before the check now pending (a run before a recheck, the stream's history) writes nothing (0.804865ms) ✔ an old run's verdict redelivered after a recheck's pending is not written over it (1.128977ms) [gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check the forge holds pending; not written [gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check now pending; not written ✔ an old run's verdict in flight when the recheck comes lands before the recheck's pending (3.056779ms) ✔ a rechecked head says where its new run is: building on which machine (1.184608ms) ✔ the recheck's own verdict is written, even asked a little before the pending by another machine's clock (0.832626ms) ✔ after a restart, an old run's verdict is weighed against the check the forge's history holds open (0.665614ms) ✔ a recheck the forge does not take fails its caller, and the head is as it was (0.598733ms) ✔ the status tool writes the merge check's pending through the writer, and the mesh's other statuses as they were (5.590877ms) ✔ when a check was asked is read from its id; the opening of the check now pending from the gate's history (0.408703ms) ✔ a head this process did not announce (it restarted) is said only while the forge holds its gate pending (1.52964ms) ✔ a repository as the mesh clones it is told from another (0.361912ms) ✔ gitea_api refuses the forge's administration, a branch's protection and the merge check's status, and sends nothing (6.1449ms) ✔ gitea_api refuses a file written to the default branch or one a rule guards, and writes one on a branch of its own (55.81583ms) ✔ a protection rule covers a branch as the forge decides it: classes, alternatives, a plain name in any case (issue 539) (30.288405ms) ✔ gitea_api still writes what an author writes (9.915634ms) ✔ gitea_branch_protection_set only strengthens (20.1742ms) ✔ gitea_merge_pull_request merges only a head whose required statuses all succeeded, and only that head (50.631746ms) ✔ gitea_merge_pull_request refuses a verdict judged against an earlier trunk, and asks the check once (150.49121ms) ✔ gitea_merge_pull_request merges only the head it is named, and refuses one that is not the head (17.699374ms) ✔ gitea_merge_pull_request is loud when the forge merged a head other than the one it was handed (13.628448ms) ✔ gitea_merge_pull_request says the merge happened when reading back what landed fails (25.294584ms) ✔ gitea_merge_pull_request reads back a fast-forward, and says it cannot read back a squash or a rebase (57.836283ms) ✔ gitea_delete_repo deletes only the tools' own user's repository, unguarded, untouched for 30 days, named in full (21.050976ms) ✔ gitea_api sends the path it judged, and refuses moving a pull request's base (3.622374ms) ✔ gitea_merge_pull_request merges only into a branch whose protection requires the checks (7.291017ms) ✔ a rule named for a branch is not made weaker than the glob rule that guards it now (13.722928ms) ✔ a repository owner or name is never another path (0.309732ms) ✔ the judges, without a forge (0.381263ms) ✔ main's protection is never weakened in its push or its merge whitelist (issue 948, ADR 0398 §2) (0.248002ms) ✔ gitea_api never acts as another user: sudo, in any case, and any query key it does not send are refused (1.783772ms) ✔ gitea_api never approves a pull request nor requests changes: those are the operator's (2.503066ms) ✔ no tool's caller text becomes another path or another user: `..` and `?sudo=` are refused before the forge (5.113573ms) ✔ gitea_merge_pull_request merges a verdict that read siblings only with its group passed and each sibling head reviewed (89.879421ms) ✔ a verdict put out of date by a sibling's move is refused in its own words, and asked again once (23.793683ms) ✔ every page of a pull request's files is read, though the forge caps a page at fifty (39.192863ms) ✔ a pass asks only the repositories that moved since the last look, every one before the first (0.614393ms) ✔ the directories above a merge's files that hold a module at the commit are said, the root never (18.087636ms) ✔ a lookup the forge refuses is thrown, not read as no module (5.740277ms) ✔ a renamed file is said under its old name too, as gone (novox/hq ADR 0267) (9.399361ms) ✔ a new head beyond the newest 20 open pull requests is announced, exactly once (75.930099ms) ✔ a pass pages until it reaches back past its floor, and no further (31.20745ms) ✔ gitea_branch_protection_get answers every rule, read with the administrator account the tools' token cannot read with (29.149917ms) ✔ gitea_branch_protection_get answers the rule for a branch, null when it has none (10.188606ms) ✔ gitea_branch_protection_set still gets the forge's refusal, as the tools' account (4.105636ms) ✔ the statuses asked for replace the rule's, and an empty list requires none (5.421895ms) ✔ a new rule refuses direct pushes unless asked; an administrator's override only when said (0.394442ms) ✔ main's body always refuses pushes and lets only the tools' account merge, whatever is asked (0.858655ms) ✔ another branch's body carries no merge whitelist (0.287392ms) ✔ the rule main is put right in the forge's own words: the body a setup sends carries no status checks of its own (0.394792ms) ✔ each open pull request's new head is announced once, and nothing closed or merged (20.358261ms) ✔ a verdict is the head commit's status; an error is the forge's error, never a success (0.955257ms) ✔ each layer is its own status: the gate with the modules it judged, the repository's own check beside it (1.013266ms) ✔ no verdict sets warning on a required check; only a person's decision fails (0.922646ms) ✔ a commit status is set on the commit, under the merge check's context (43.049537ms) ✔ a change plan is the gate's result: said on the status and, when it builds something, as a comment (0.571763ms) ✔ a head announced is pending in the build queue, since the local time it was announced (0.350792ms) ✔ the vault's token is used, a setting of it is ignored and said once, without its value (2.573667ms) [gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 at bbbbbbbbbbbb, it is now at dddddddddddd: out of date, nothing run (novox/hq ADR 0382) [gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was closed unmerged: out of date, nothing run (novox/hq ADR 0382) [gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 merged onto cccccccccccc, its main is now eeeeeeeeeeee: out of date, nothing run (novox/hq ADR 0382) [gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was merged otherwise than its head bbbbbbbbbbbb onto cccccccccccc: out of date, nothing run (novox/hq ADR 0382) [gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was merged otherwise than its head bbbbbbbbbbbb onto cccccccccccc: out of date, nothing run (novox/hq ADR 0382) [gitea] novox/mesh-host's main moved from 444444444444 to 333333333333: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323) ✔ both statuses name the siblings the verdict read, keeping the verdict word and the start of its reason within 140 characters (2.593177ms) ✔ a verdict that could not learn its siblings says so on the status and in the comment, and names none (0.428613ms) ✔ a verdict without siblings is said exactly as before (an older controller) (0.856436ms) ✔ the siblings a verdict names are kept per member, replaced by its next verdict, and removed by one that read none (1.009537ms) ✔ a sibling's new head puts both statuses of the member back to pending, with the reason, and runs nothing (3.916835ms) ✔ a sibling closed unmerged puts the member's verdict out of date (0.503023ms) ✔ a move of the sibling's base puts it out of date, naming both commits (0.508053ms) ✔ the sibling's own merge of the head read onto the base read does not, in either order of the two merges (0.690205ms) ✔ a merge of another head, or onto another base, is a move (1.148408ms) ✔ only the members whose verdict names the sibling are touched, and a member that moved on is dropped (0.850135ms) ✔ a trunk that moved tells the siblings' weighing; one first seen does not (0.859435ms) ✔ the group rule names every missing thing: the group status at each head read, and each sibling head's review (1.073487ms) ✔ a merged or closed member's entry is dropped, so a sibling's later push never re-pends its merged head (0.270672ms) ✔ a consumer that asks nothing, or write, publishes: the package team (2.224155ms) ✔ a consumer that asks to read joins the readers, who write no package (0.720095ms) ✔ a reader is made, put on the readers team, and taken off the package team (0.549103ms) ✔ a publisher is made and put on the package team, as before (0.224721ms) ✔ first start: mints with the admin account, keeps the token at 0600, asks for the tools' scopes only (44.501937ms) ✔ second start: reuses the kept token, mints nothing (28.982286ms) ✔ the forge rejects the kept token (its data was restored): minted afresh, once, and the call goes through (28.976037ms) ✔ a token kept with write:admin by an earlier build is minted again without it at the next start, and the forge drops the old one (32.020797ms) ✔ a kept token lacking a scope the tools need: the forge refuses the call for the scope, the token is re-minted with the whole list, and the call goes through (27.696079ms) ✔ the kept file is gone but the forge still holds a token by that name: replaced, not refused (25.309174ms) ✔ concurrent first calls share one mint (16.507176ms) ✔ two tokens hold the name (two mints raced): both are deleted by their id, and one fresh token is minted (16.284155ms) ✔ two processes on one kept file, both without a token, mint once: the second waits for the first and reuses it (62.02393ms) ✔ a lock left by a process that died mid-mint is taken over, not waited on for ever (6.028509ms) ✔ the admin account is refused: said plainly, nothing kept, and the next call fails the same way rather than crashing (17.449353ms) ✔ one process shares one source per kept file — the watcher and the tools never renew against each other (0.814335ms) ✔ a second process finds the token the first renewed, and reuses it instead of minting over it (22.728606ms) ✔ a configured token wins, and is reported rather than minted over when the forge rejects it (2.704307ms) ✔ nothing to mint with and no token: the client says what is missing (0.213322ms) [gitea] no tools — no Gitea token, and nothing to mint one with — set MESH_GITEA_TOKEN, or MESH_GITEA_ADMIN_USER, MESH_GITEA_ADMIN_PASSWORD_FILE, MESH_GITEA_STATE_DIR [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-dT9Z6L/state/token ✔ the tools register once there is a way to a token, and the first call mints it (29.48841ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-pR2QPr/state/token ✔ a pull request can be closed, read and cleaned up through the tools (37.673513ms) [gitea] no token kept — minting one [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-V03qOP/state/token ✔ the tools' token is minted for an account of their own, which is no administrator and writes through its team (93.23407ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-6UD3o4/state/token ✔ with the tools' token, the forge writes a status and refuses a protection edit — the forge itself, not a guard (62.276681ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-2Ym0YR/state/token ✔ what a base's protection asks is read from the branch, which the tools' account may read, not from the rules' list it may not (50.752207ms) [gitea] the kept token was minted with other scopes than the tools take now — minting one with them [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-aXncVG/state/token [gitea] deleted 1 token(s) named "mesh-tools" the admin account "mesh-admin" held: the tools act as "mesh-tools" now ✔ the administrator's token from before goes: a kept admin token is minted again for the tools' account, and the forge drops the old one (52.544168ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-S7FEmb/state/token ✔ a tools' account someone made an administrator is made a plain account again at the next mint (38.297558ms) ✔ the tools never act as the administrator: MESH_GITEA_TOOLS_USER naming it is refused (1.012227ms) ✔ the manifest hands the tools their account, and the merge-ready authors name it rather than the administrator (1.042757ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-RH9eMW/state/token ✔ a repository check said as a warning on a head without merge-check.sh reads the base's protection from the branch (41.591149ms) [gitea] no token kept — minting one ✔ a mint that fails leaves the administrator's old token in place: the tools are never left without one (31.326852ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-u4A0Fu/state/token [gitea] FAILED to delete the token(s) named "mesh-tools" the admin account "mesh-admin" holds — tried again at the next start: Gitea DELETE /users/mesh-admin/tokens/902 (named mesh-tools): 500 {"message":"the forge is not well"} ✔ an old token that cannot be deleted after the mint is said, the new token kept, and deleted at the next start (49.043327ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-eVEpdU/state/token ✔ a team another process made between the look and the create is found, not a failure (42.530485ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-z9RksA/state/token ✔ setting the tools' account up puts main's rule right in every repository: no push, only the tools' account merges (issue 948) (38.526369ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-Dnd9F7/state/token ✔ a glob rule covering main keeps its statuses when main is given a rule of its own name (issue 948, review round 1) (38.187117ms) ✔ a start with the token kept still puts main's rules right: the module's start asks for it each time (issue 948, review round 1) (13.581238ms) [gitea] novox/mesh-host's main moved from a2c2c35bbbbb to d28ca95ccccc: 25 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323) [gitea] novox/mesh-host's main moved from d28ca95ccccc to ffffffffffff: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323) [gitea] novox/mesh-host: whether a trunk moved could not be weighed, so it is weighed again on the next pass — the forge is away [gitea] novox/mesh-host's main moved from a2c2c35bbbbb to d28ca95ccccc: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323) [gitea] novox/mesh-controller's feat/57-config-and-provisioning moved from eeeeeeeeeeee to d28ca95ccccc: 1 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323) ✔ both statuses say the head and the trunk commit the verdict judged, first, whatever the summary's length (2.495656ms) ✔ a trunk's move puts an open head's merge check out of date, naming both commits, unless a check is on its way (1.4566ms) ✔ a push to a trunk puts every open pull request into it out of date — twenty-five of them — and none into another (5.841468ms) ✔ a trunk that cannot be read is weighed again on the next pass, never recorded half (0.951097ms) ✔ the trunk's move takes its turn after a verdict in flight, and weighs what the forge holds then (0.585294ms) ✔ a verdict judged against a trunk newer than the pass's first read is not put out of date (0.345292ms) ✔ a verdict that landed out of date keeps its account on the pull request (0.446332ms) ✔ a release branch's move puts every open pull request into it out of date, as a trunk's does (ADR 0358 §5) (1.464999ms) ✔ a signed event is published on the bus and acted on once (67.290554ms) ✔ an unsigned event, or one whose signature does not verify, is refused and published nowhere (39.054402ms) ✔ an event and the poll seeing the same head announce it once (0.563423ms) ✔ each forge event is named in the module's stream and trimmed (0.735795ms) ✔ the webhook is made a system webhook, which the forge sends every repository's events to (1.47756ms) ✔ default webhooks left at the receiver by an earlier start are replaced by one system webhook (1.428569ms) ✔ a repository's copy of a default webhook at the receiver is deleted, on every page, and said (5.287073ms) ✔ a forge that makes no system webhook is said, never taken as heard (0.993257ms) ✔ the receiver tells what it verified and what it refused (10.845129ms) ✔ a forge that never delivers is said loudly once, and unhealthy, when the poll finds what no event said (1.365349ms) ✔ a forge whose events stop after one delivery is said loudly once per silent spell, and refusals are unhealthy (0.532724ms) ✔ without a receiver by configuration the polls alone watch, healthy; a receiver that failed is unhealthy (0.434543ms) ✔ the polls look every 30 s until events are heard, every five minutes while they are, and 30 s again on a miss (0.258161ms) ℹ tests 147 ℹ suites 0 ℹ pass 147 ℹ fail 0 ℹ cancelled 0 ℹ skipped 0 ℹ todo 0 ℹ duration_ms 849.901758 ```

mesh-tools added 1 commit 2026-10-16 22:05:53 +00:00

Hand ssh-client's forge files to the provisioner client (issue 955) ae9cda8531

mesh-tools commented 2026-10-16 22:08:32 +00:00

Author

Member

Copy Link

Copy Source

Merge check at ae9cda85
Judged: head ae9cda8531b3 merged into trunk 017b729bc95d; read beside it: mesh-controller 6a62a7e552ea, mesh-controller-main 6a62a7e552ea, mesh-host cd13f7726ef0, mesh-tools 7412c5b6c9c5

• mesh/merge-gate: PASS — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea, ssh-client)
• mesh/repo-check: PASS — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript)

Delivery plan — builds gitea, ssh-client → ace, g14, novox, shanks; no bus step

• tier 0: gitea, ssh-client
• ace: receives ssh-client
• g14: receives ssh-client
• novox: receives gitea, ssh-client
• shanks: receives ssh-client
• gitea provides npm-package-registry, git, gitea-read: its consumers are sent again after it
• gitea keeps data (ADR 0233): what it holds is backed up before it moves
• ssh-client keeps data (ADR 0233): what it holds is backed up before it moves
• grants step: if this changes what the bus's user list says (a new tool, subject or user), novox is sent that list alone, every build there kept, before ssh-client is judged on its first machine
• read by no module's build: modules/ssh-client/cmd/ssh-client-tools/keykeeper_bus_test.go, modules/ssh-client/cmd/ssh-client-tools/keykeeper_test.go, modules/ssh-client/cmd/ssh-client-tools/templates_test.go

Run by the build seat on ace as build-1792188355270468326 (builds --log build-1792188355270468326).

✔ the accounts that may mark are the setting's, and only the mesh's own when it cannot be read (0.837457ms)
✔ the newest status of each context is kept, with what it said, when and by whom (27.33629ms)
✔ a commit merges only when every required status is a success; a warning blocks (0.70645ms)
✔ no rule requires nothing; a rule that cannot be read says nothing, never yes (0.293242ms)
✔ a commit is contained in another when the compare finds nothing on it the other lacks (3.279455ms)
✔ a note line is appended once, as the forge's user, in its own repository (9.826752ms)
✔ a note that cannot be written is said, never read as written (0.374956ms)
✔ one view per pull request, found by its marker; only the mesh's statuses (0.505021ms)
✔ a build machine taking the check is said on the head: building on which machine, since the local time (19.394717ms)
✔ the verdict heard while a 'building' write is in flight lands last and stays (5.007638ms)
[gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check now pending; not written
✔ the verdict written while a start's read of the forge is in flight stays the newest status (0.816768ms)
✔ the verdict heard while a 'building' write still waits its turn: the 'building' is never written (0.403239ms)
✔ a start heard after the verdict writes nothing (0.380046ms)
✔ a start from before the check now pending (a run before a recheck, the stream's history) writes nothing (0.45158ms)
✔ an old run's verdict redelivered after a recheck's pending is not written over it (0.681524ms)
[gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check the forge holds pending; not written
[gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check now pending; not written
✔ an old run's verdict in flight when the recheck comes lands before the recheck's pending (1.472322ms)
✔ a rechecked head says where its new run is: building on which machine (0.582597ms)
✔ the recheck's own verdict is written, even asked a little before the pending by another machine's clock (0.625958ms)
✔ after a restart, an old run's verdict is weighed against the check the forge's history holds open (0.406816ms)
✔ a recheck the forge does not take fails its caller, and the head is as it was (0.381058ms)
✔ the status tool writes the merge check's pending through the writer, and the mesh's other statuses as they were (4.598768ms)
✔ when a check was asked is read from its id; the opening of the check now pending from the gate's history (0.263265ms)
✔ a head this process did not announce (it restarted) is said only while the forge holds its gate pending (3.490441ms)
✔ a repository as the mesh clones it is told from another (0.300256ms)
✔ gitea_api refuses the forge's administration, a branch's protection and the merge check's status, and sends nothing (6.16096ms)
✔ gitea_api refuses a file written to the default branch or one a rule guards, and writes one on a branch of its own (49.20173ms)
✔ a protection rule covers a branch as the forge decides it: classes, alternatives, a plain name in any case (issue 539) (21.621479ms)
✔ gitea_api still writes what an author writes (7.22888ms)
✔ gitea_branch_protection_set only strengthens (15.910648ms)
✔ gitea_merge_pull_request merges only a head whose required statuses all succeeded, and only that head (46.303702ms)
✔ gitea_merge_pull_request refuses a verdict judged against an earlier trunk, and asks the check once (152.22756ms)
✔ gitea_merge_pull_request merges only the head it is named, and refuses one that is not the head (17.530587ms)
✔ gitea_merge_pull_request is loud when the forge merged a head other than the one it was handed (11.115259ms)
✔ gitea_merge_pull_request says the merge happened when reading back what landed fails (25.213382ms)
✔ gitea_merge_pull_request reads back a fast-forward, and says it cannot read back a squash or a rebase (55.950085ms)
✔ gitea_delete_repo deletes only the tools' own user's repository, unguarded, untouched for 30 days, named in full (21.62735ms)
✔ gitea_api sends the path it judged, and refuses moving a pull request's base (3.93563ms)
✔ gitea_merge_pull_request merges only into a branch whose protection requires the checks (6.08646ms)
✔ a rule named for a branch is not made weaker than the glob rule that guards it now (14.221748ms)
✔ a repository owner or name is never another path (0.216357ms)
✔ the judges, without a forge (0.351722ms)
✔ main's protection is never weakened in its push or its merge whitelist (issue 948, ADR 0398 §2) (0.231065ms)
✔ gitea_api never acts as another user: sudo, in any case, and any query key it does not send are refused (2.399949ms)
✔ gitea_api never approves a pull request nor requests changes: those are the operator's (1.218965ms)
✔ no tool's caller text becomes another path or another user: `..` and `?sudo=` are refused before the forge (4.534938ms)
✔ gitea_merge_pull_request merges a verdict that read siblings only with its group passed and each sibling head reviewed (79.695274ms)
✔ a verdict put out of date by a sibling's move is refused in its own words, and asked again once (23.101406ms)
✔ every page of a pull request's files is read, though the forge caps a page at fifty (25.81226ms)
✔ a pass asks only the repositories that moved since the last look, every one before the first (0.608005ms)
✔ the directories above a merge's files that hold a module at the commit are said, the root never (13.623351ms)
✔ a lookup the forge refuses is thrown, not read as no module (4.554185ms)
✔ a renamed file is said under its old name too, as gone (novox/hq ADR 0267) (6.472246ms)
✔ a new head beyond the newest 20 open pull requests is announced, exactly once (57.27486ms)
✔ a pass pages until it reaches back past its floor, and no further (18.338178ms)
✔ gitea_branch_protection_get answers every rule, read with the administrator account the tools' token cannot read with (24.685128ms)
✔ gitea_branch_protection_get answers the rule for a branch, null when it has none (5.954061ms)
✔ gitea_branch_protection_set still gets the forge's refusal, as the tools' account (3.029203ms)
✔ the statuses asked for replace the rule's, and an empty list requires none (4.267545ms)
✔ a new rule refuses direct pushes unless asked; an administrator's override only when said (0.319021ms)
✔ main's body always refuses pushes and lets only the tools' account merge, whatever is asked (0.793344ms)
✔ another branch's body carries no merge whitelist (0.252385ms)
✔ the rule main is put right in the forge's own words: the body a setup sends carries no status checks of its own (0.335622ms)
✔ each open pull request's new head is announced once, and nothing closed or merged (14.618004ms)
✔ a verdict is the head commit's status; an error is the forge's error, never a success (1.034728ms)
✔ each layer is its own status: the gate with the modules it judged, the repository's own check beside it (0.704687ms)
✔ no verdict sets warning on a required check; only a person's decision fails (1.042683ms)
✔ a commit status is set on the commit, under the merge check's context (40.63518ms)
✔ a change plan is the gate's result: said on the status and, when it builds something, as a comment (0.569222ms)
✔ a head announced is pending in the build queue, since the local time it was announced (0.369105ms)
✔ the vault's token is used, a setting of it is ignored and said once, without its value (3.009777ms)
[gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 at bbbbbbbbbbbb, it is now at dddddddddddd: out of date, nothing run (novox/hq ADR 0382)
[gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was closed unmerged: out of date, nothing run (novox/hq ADR 0382)
[gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 merged onto cccccccccccc, its main is now eeeeeeeeeeee: out of date, nothing run (novox/hq ADR 0382)
[gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was merged otherwise than its head bbbbbbbbbbbb onto cccccccccccc: out of date, nothing run (novox/hq ADR 0382)
[gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was merged otherwise than its head bbbbbbbbbbbb onto cccccccccccc: out of date, nothing run (novox/hq ADR 0382)
[gitea] novox/mesh-host's main moved from 444444444444 to 333333333333: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323)
✔ both statuses name the siblings the verdict read, keeping the verdict word and the start of its reason within 140 characters (2.115433ms)
✔ a verdict that could not learn its siblings says so on the status and in the comment, and names none (0.319161ms)
✔ a verdict without siblings is said exactly as before (an older controller) (0.740134ms)
✔ the siblings a verdict names are kept per member, replaced by its next verdict, and removed by one that read none (0.890507ms)
✔ a sibling's new head puts both statuses of the member back to pending, with the reason, and runs nothing (7.392118ms)
✔ a sibling closed unmerged puts the member's verdict out of date (0.411285ms)
✔ a move of the sibling's base puts it out of date, naming both commits (0.379083ms)
✔ the sibling's own merge of the head read onto the base read does not, in either order of the two merges (0.620157ms)
✔ a merge of another head, or onto another base, is a move (0.771583ms)
✔ only the members whose verdict names the sibling are touched, and a member that moved on is dropped (1.823874ms)
✔ a trunk that moved tells the siblings' weighing; one first seen does not (0.683487ms)
✔ the group rule names every missing thing: the group status at each head read, and each sibling head's review (1.00898ms)
✔ a merged or closed member's entry is dropped, so a sibling's later push never re-pends its merged head (0.268997ms)
✔ a consumer that asks nothing, or write, publishes: the package team (1.125389ms)
✔ a consumer that asks to read joins the readers, who write no package (0.369696ms)
✔ a reader is made, put on the readers team, and taken off the package team (0.542031ms)
✔ a publisher is made and put on the package team, as before (0.160812ms)
✔ first start: mints with the admin account, keeps the token at 0600, asks for the tools' scopes only (38.699715ms)
✔ second start: reuses the kept token, mints nothing (23.396783ms)
✔ the forge rejects the kept token (its data was restored): minted afresh, once, and the call goes through (22.307321ms)
✔ a token kept with write:admin by an earlier build is minted again without it at the next start, and the forge drops the old one (25.942255ms)
✔ a kept token lacking a scope the tools need: the forge refuses the call for the scope, the token is re-minted with the whole list, and the call goes through (25.438236ms)
✔ the kept file is gone but the forge still holds a token by that name: replaced, not refused (22.747009ms)
✔ concurrent first calls share one mint (11.037461ms)
✔ two tokens hold the name (two mints raced): both are deleted by their id, and one fresh token is minted (13.03344ms)
✔ two processes on one kept file, both without a token, mint once: the second waits for the first and reuses it (58.902484ms)
✔ a lock left by a process that died mid-mint is taken over, not waited on for ever (6.852602ms)
✔ the admin account is refused: said plainly, nothing kept, and the next call fails the same way rather than crashing (16.569057ms)
✔ one process shares one source per kept file — the watcher and the tools never renew against each other (0.740084ms)
✔ a second process finds the token the first renewed, and reuses it instead of minting over it (19.470681ms)
✔ a configured token wins, and is reported rather than minted over when the forge rejects it (2.313185ms)
✔ nothing to mint with and no token: the client says what is missing (0.189216ms)
[gitea] no tools — no Gitea token, and nothing to mint one with — set MESH_GITEA_TOKEN, or MESH_GITEA_ADMIN_USER, MESH_GITEA_ADMIN_PASSWORD_FILE, MESH_GITEA_STATE_DIR
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-g4HfEx/state/token
✔ the tools register once there is a way to a token, and the first call mints it (26.93814ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-1Qqrag/state/token
✔ a pull request can be closed, read and cleaned up through the tools (36.589723ms)
[gitea] no token kept — minting one
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-ByUaYi/state/token
✔ the tools' token is minted for an account of their own, which is no administrator and writes through its team (70.799354ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-M3zXJl/state/token
✔ with the tools' token, the forge writes a status and refuses a protection edit — the forge itself, not a guard (48.430046ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-0qm1CV/state/token
✔ what a base's protection asks is read from the branch, which the tools' account may read, not from the rules' list it may not (49.305635ms)
[gitea] the kept token was minted with other scopes than the tools take now — minting one with them
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-FJ9Q2j/state/token
[gitea] deleted 1 token(s) named "mesh-tools" the admin account "mesh-admin" held: the tools act as "mesh-tools" now
✔ the administrator's token from before goes: a kept admin token is minted again for the tools' account, and the forge drops the old one (42.872422ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-TwIXjX/state/token
✔ a tools' account someone made an administrator is made a plain account again at the next mint (36.937157ms)
✔ the tools never act as the administrator: MESH_GITEA_TOOLS_USER naming it is refused (0.8763ms)
✔ the manifest hands the tools their account, and the merge-ready authors name it rather than the administrator (1.044717ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-Cpae4b/state/token
✔ a repository check said as a warning on a head without merge-check.sh reads the base's protection from the branch (37.45392ms)
[gitea] no token kept — minting one
✔ a mint that fails leaves the administrator's old token in place: the tools are never left without one (27.773352ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-zmQCEc/state/token
[gitea] FAILED to delete the token(s) named "mesh-tools" the admin account "mesh-admin" holds — tried again at the next start: Gitea DELETE /users/mesh-admin/tokens/902 (named mesh-tools): 500 {"message":"the forge is not well"}
✔ an old token that cannot be deleted after the mint is said, the new token kept, and deleted at the next start (43.762187ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-h2lcrr/state/token
✔ a team another process made between the look and the create is found, not a failure (38.936211ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-p8E6kK/state/token
✔ setting the tools' account up puts main's rule right in every repository: no push, only the tools' account merges (issue 948) (37.47518ms)
[gitea] no token kept — minting one
[gitea] the forge already holds a token named "mesh-tools" — replacing it
[gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-Vk5wtD/state/token
✔ a glob rule covering main keeps its statuses when main is given a rule of its own name (issue 948, review round 1) (35.260179ms)
✔ a start with the token kept still puts main's rules right: the module's start asks for it each time (issue 948, review round 1) (13.215432ms)
[gitea] novox/mesh-host's main moved from a2c2c35bbbbb to d28ca95ccccc: 25 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323)
[gitea] novox/mesh-host's main moved from d28ca95ccccc to ffffffffffff: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323)
[gitea] novox/mesh-host: whether a trunk moved could not be weighed, so it is weighed again on the next pass — the forge is away
[gitea] novox/mesh-host's main moved from a2c2c35bbbbb to d28ca95ccccc: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323)
[gitea] novox/mesh-controller's feat/57-config-and-provisioning moved from eeeeeeeeeeee to d28ca95ccccc: 1 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323)
✔ both statuses say the head and the trunk commit the verdict judged, first, whatever the summary's length (1.48711ms)
✔ a trunk's move puts an open head's merge check out of date, naming both commits, unless a check is on its way (0.806289ms)
✔ a push to a trunk puts every open pull request into it out of date — twenty-five of them — and none into another (2.970192ms)
✔ a trunk that cannot be read is weighed again on the next pass, never recorded half (0.287672ms)
✔ the trunk's move takes its turn after a verdict in flight, and weighs what the forge holds then (0.385335ms)
✔ a verdict judged against a trunk newer than the pass's first read is not put out of date (0.231306ms)
✔ a verdict that landed out of date keeps its account on the pull request (0.213181ms)
✔ a release branch's move puts every open pull request into it out of date, as a trunk's does (ADR 0358 §5) (0.408589ms)
✔ a signed event is published on the bus and acted on once (52.481775ms)
✔ an unsigned event, or one whose signature does not verify, is refused and published nowhere (31.605187ms)
✔ an event and the poll seeing the same head announce it once (0.305576ms)
✔ each forge event is named in the module's stream and trimmed (0.46733ms)
✔ the webhook is made a system webhook, which the forge sends every repository's events to (1.113676ms)
✔ default webhooks left at the receiver by an earlier start are replaced by one system webhook (0.990175ms)
✔ a repository's copy of a default webhook at the receiver is deleted, on every page, and said (3.078197ms)
✔ a forge that makes no system webhook is said, never taken as heard (0.675772ms)
✔ the receiver tells what it verified and what it refused (7.665953ms)
✔ a forge that never delivers is said loudly once, and unhealthy, when the poll finds what no event said (1.20583ms)
✔ a forge whose events stop after one delivery is said loudly once per silent spell, and refusals are unhealthy (0.724263ms)
✔ without a receiver by configuration the polls alone watch, healthy; a receiver that failed is unhealthy (0.343116ms)
✔ the polls look every 30 s until events are heard, every five minutes while they are, and 30 s again on a miss (0.187533ms)
ℹ tests 147
ℹ suites 0
ℹ pass 147
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 751.836759

**Merge check** at `ae9cda85` Judged: head `ae9cda8531b3` merged into trunk `017b729bc95d`; read beside it: mesh-controller `6a62a7e552ea`, mesh-controller-main `6a62a7e552ea`, mesh-host `cd13f7726ef0`, mesh-tools `7412c5b6c9c5` - `mesh/merge-gate`: **PASS** — every machine composes with the change as it did without (4 of 4 compose) (modules: gitea, ssh-client) - `mesh/repo-check`: **PASS** — its merge-check.sh passed, each part in its toolchain: merge-check.sh (go), merge-check-typescript.sh (typescript) **Delivery plan** — builds gitea, ssh-client → ace, g14, novox, shanks; no bus step - tier 0: gitea, ssh-client - ace: receives ssh-client - g14: receives ssh-client - novox: receives gitea, ssh-client - shanks: receives ssh-client - gitea provides npm-package-registry, git, gitea-read: its consumers are sent again after it - gitea keeps data (ADR 0233): what it holds is backed up before it moves - ssh-client keeps data (ADR 0233): what it holds is backed up before it moves - grants step: if this changes what the bus's user list says (a new tool, subject or user), novox is sent that list alone, every build there kept, before ssh-client is judged on its first machine - read by no module's build: modules/ssh-client/cmd/ssh-client-tools/keykeeper_bus_test.go, modules/ssh-client/cmd/ssh-client-tools/keykeeper_test.go, modules/ssh-client/cmd/ssh-client-tools/templates_test.go Run by the build seat on ace as `build-1792188355270468326` (`builds --log build-1792188355270468326`). ``` ✔ the accounts that may mark are the setting's, and only the mesh's own when it cannot be read (0.837457ms) ✔ the newest status of each context is kept, with what it said, when and by whom (27.33629ms) ✔ a commit merges only when every required status is a success; a warning blocks (0.70645ms) ✔ no rule requires nothing; a rule that cannot be read says nothing, never yes (0.293242ms) ✔ a commit is contained in another when the compare finds nothing on it the other lacks (3.279455ms) ✔ a note line is appended once, as the forge's user, in its own repository (9.826752ms) ✔ a note that cannot be written is said, never read as written (0.374956ms) ✔ one view per pull request, found by its marker; only the mesh's statuses (0.505021ms) ✔ a build machine taking the check is said on the head: building on which machine, since the local time (19.394717ms) ✔ the verdict heard while a 'building' write is in flight lands last and stays (5.007638ms) [gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check now pending; not written ✔ the verdict written while a start's read of the forge is in flight stays the newest status (0.816768ms) ✔ the verdict heard while a 'building' write still waits its turn: the 'building' is never written (0.403239ms) ✔ a start heard after the verdict writes nothing (0.380046ms) ✔ a start from before the check now pending (a run before a recheck, the stream's history) writes nothing (0.45158ms) ✔ an old run's verdict redelivered after a recheck's pending is not written over it (0.681524ms) [gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check the forge holds pending; not written [gitea] novox/hq@a6b08de8: the verdict of build-1791680941000000000 is an older run's than the check now pending; not written ✔ an old run's verdict in flight when the recheck comes lands before the recheck's pending (1.472322ms) ✔ a rechecked head says where its new run is: building on which machine (0.582597ms) ✔ the recheck's own verdict is written, even asked a little before the pending by another machine's clock (0.625958ms) ✔ after a restart, an old run's verdict is weighed against the check the forge's history holds open (0.406816ms) ✔ a recheck the forge does not take fails its caller, and the head is as it was (0.381058ms) ✔ the status tool writes the merge check's pending through the writer, and the mesh's other statuses as they were (4.598768ms) ✔ when a check was asked is read from its id; the opening of the check now pending from the gate's history (0.263265ms) ✔ a head this process did not announce (it restarted) is said only while the forge holds its gate pending (3.490441ms) ✔ a repository as the mesh clones it is told from another (0.300256ms) ✔ gitea_api refuses the forge's administration, a branch's protection and the merge check's status, and sends nothing (6.16096ms) ✔ gitea_api refuses a file written to the default branch or one a rule guards, and writes one on a branch of its own (49.20173ms) ✔ a protection rule covers a branch as the forge decides it: classes, alternatives, a plain name in any case (issue 539) (21.621479ms) ✔ gitea_api still writes what an author writes (7.22888ms) ✔ gitea_branch_protection_set only strengthens (15.910648ms) ✔ gitea_merge_pull_request merges only a head whose required statuses all succeeded, and only that head (46.303702ms) ✔ gitea_merge_pull_request refuses a verdict judged against an earlier trunk, and asks the check once (152.22756ms) ✔ gitea_merge_pull_request merges only the head it is named, and refuses one that is not the head (17.530587ms) ✔ gitea_merge_pull_request is loud when the forge merged a head other than the one it was handed (11.115259ms) ✔ gitea_merge_pull_request says the merge happened when reading back what landed fails (25.213382ms) ✔ gitea_merge_pull_request reads back a fast-forward, and says it cannot read back a squash or a rebase (55.950085ms) ✔ gitea_delete_repo deletes only the tools' own user's repository, unguarded, untouched for 30 days, named in full (21.62735ms) ✔ gitea_api sends the path it judged, and refuses moving a pull request's base (3.93563ms) ✔ gitea_merge_pull_request merges only into a branch whose protection requires the checks (6.08646ms) ✔ a rule named for a branch is not made weaker than the glob rule that guards it now (14.221748ms) ✔ a repository owner or name is never another path (0.216357ms) ✔ the judges, without a forge (0.351722ms) ✔ main's protection is never weakened in its push or its merge whitelist (issue 948, ADR 0398 §2) (0.231065ms) ✔ gitea_api never acts as another user: sudo, in any case, and any query key it does not send are refused (2.399949ms) ✔ gitea_api never approves a pull request nor requests changes: those are the operator's (1.218965ms) ✔ no tool's caller text becomes another path or another user: `..` and `?sudo=` are refused before the forge (4.534938ms) ✔ gitea_merge_pull_request merges a verdict that read siblings only with its group passed and each sibling head reviewed (79.695274ms) ✔ a verdict put out of date by a sibling's move is refused in its own words, and asked again once (23.101406ms) ✔ every page of a pull request's files is read, though the forge caps a page at fifty (25.81226ms) ✔ a pass asks only the repositories that moved since the last look, every one before the first (0.608005ms) ✔ the directories above a merge's files that hold a module at the commit are said, the root never (13.623351ms) ✔ a lookup the forge refuses is thrown, not read as no module (4.554185ms) ✔ a renamed file is said under its old name too, as gone (novox/hq ADR 0267) (6.472246ms) ✔ a new head beyond the newest 20 open pull requests is announced, exactly once (57.27486ms) ✔ a pass pages until it reaches back past its floor, and no further (18.338178ms) ✔ gitea_branch_protection_get answers every rule, read with the administrator account the tools' token cannot read with (24.685128ms) ✔ gitea_branch_protection_get answers the rule for a branch, null when it has none (5.954061ms) ✔ gitea_branch_protection_set still gets the forge's refusal, as the tools' account (3.029203ms) ✔ the statuses asked for replace the rule's, and an empty list requires none (4.267545ms) ✔ a new rule refuses direct pushes unless asked; an administrator's override only when said (0.319021ms) ✔ main's body always refuses pushes and lets only the tools' account merge, whatever is asked (0.793344ms) ✔ another branch's body carries no merge whitelist (0.252385ms) ✔ the rule main is put right in the forge's own words: the body a setup sends carries no status checks of its own (0.335622ms) ✔ each open pull request's new head is announced once, and nothing closed or merged (14.618004ms) ✔ a verdict is the head commit's status; an error is the forge's error, never a success (1.034728ms) ✔ each layer is its own status: the gate with the modules it judged, the repository's own check beside it (0.704687ms) ✔ no verdict sets warning on a required check; only a person's decision fails (1.042683ms) ✔ a commit status is set on the commit, under the merge check's context (40.63518ms) ✔ a change plan is the gate's result: said on the status and, when it builds something, as a comment (0.569222ms) ✔ a head announced is pending in the build queue, since the local time it was announced (0.369105ms) ✔ the vault's token is used, a setting of it is ignored and said once, without its value (3.009777ms) [gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 at bbbbbbbbbbbb, it is now at dddddddddddd: out of date, nothing run (novox/hq ADR 0382) [gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was closed unmerged: out of date, nothing run (novox/hq ADR 0382) [gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 merged onto cccccccccccc, its main is now eeeeeeeeeeee: out of date, nothing run (novox/hq ADR 0382) [gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was merged otherwise than its head bbbbbbbbbbbb onto cccccccccccc: out of date, nothing run (novox/hq ADR 0382) [gitea] novox/mesh-controller#346: its verdict was judged beside mesh-host#114 was merged otherwise than its head bbbbbbbbbbbb onto cccccccccccc: out of date, nothing run (novox/hq ADR 0382) [gitea] novox/mesh-host's main moved from 444444444444 to 333333333333: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323) ✔ both statuses name the siblings the verdict read, keeping the verdict word and the start of its reason within 140 characters (2.115433ms) ✔ a verdict that could not learn its siblings says so on the status and in the comment, and names none (0.319161ms) ✔ a verdict without siblings is said exactly as before (an older controller) (0.740134ms) ✔ the siblings a verdict names are kept per member, replaced by its next verdict, and removed by one that read none (0.890507ms) ✔ a sibling's new head puts both statuses of the member back to pending, with the reason, and runs nothing (7.392118ms) ✔ a sibling closed unmerged puts the member's verdict out of date (0.411285ms) ✔ a move of the sibling's base puts it out of date, naming both commits (0.379083ms) ✔ the sibling's own merge of the head read onto the base read does not, in either order of the two merges (0.620157ms) ✔ a merge of another head, or onto another base, is a move (0.771583ms) ✔ only the members whose verdict names the sibling are touched, and a member that moved on is dropped (1.823874ms) ✔ a trunk that moved tells the siblings' weighing; one first seen does not (0.683487ms) ✔ the group rule names every missing thing: the group status at each head read, and each sibling head's review (1.00898ms) ✔ a merged or closed member's entry is dropped, so a sibling's later push never re-pends its merged head (0.268997ms) ✔ a consumer that asks nothing, or write, publishes: the package team (1.125389ms) ✔ a consumer that asks to read joins the readers, who write no package (0.369696ms) ✔ a reader is made, put on the readers team, and taken off the package team (0.542031ms) ✔ a publisher is made and put on the package team, as before (0.160812ms) ✔ first start: mints with the admin account, keeps the token at 0600, asks for the tools' scopes only (38.699715ms) ✔ second start: reuses the kept token, mints nothing (23.396783ms) ✔ the forge rejects the kept token (its data was restored): minted afresh, once, and the call goes through (22.307321ms) ✔ a token kept with write:admin by an earlier build is minted again without it at the next start, and the forge drops the old one (25.942255ms) ✔ a kept token lacking a scope the tools need: the forge refuses the call for the scope, the token is re-minted with the whole list, and the call goes through (25.438236ms) ✔ the kept file is gone but the forge still holds a token by that name: replaced, not refused (22.747009ms) ✔ concurrent first calls share one mint (11.037461ms) ✔ two tokens hold the name (two mints raced): both are deleted by their id, and one fresh token is minted (13.03344ms) ✔ two processes on one kept file, both without a token, mint once: the second waits for the first and reuses it (58.902484ms) ✔ a lock left by a process that died mid-mint is taken over, not waited on for ever (6.852602ms) ✔ the admin account is refused: said plainly, nothing kept, and the next call fails the same way rather than crashing (16.569057ms) ✔ one process shares one source per kept file — the watcher and the tools never renew against each other (0.740084ms) ✔ a second process finds the token the first renewed, and reuses it instead of minting over it (19.470681ms) ✔ a configured token wins, and is reported rather than minted over when the forge rejects it (2.313185ms) ✔ nothing to mint with and no token: the client says what is missing (0.189216ms) [gitea] no tools — no Gitea token, and nothing to mint one with — set MESH_GITEA_TOKEN, or MESH_GITEA_ADMIN_USER, MESH_GITEA_ADMIN_PASSWORD_FILE, MESH_GITEA_STATE_DIR [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-g4HfEx/state/token ✔ the tools register once there is a way to a token, and the first call mints it (26.93814ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-1Qqrag/state/token ✔ a pull request can be closed, read and cleaned up through the tools (36.589723ms) [gitea] no token kept — minting one [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-ByUaYi/state/token ✔ the tools' token is minted for an account of their own, which is no administrator and writes through its team (70.799354ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-M3zXJl/state/token ✔ with the tools' token, the forge writes a status and refuses a protection edit — the forge itself, not a guard (48.430046ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-0qm1CV/state/token ✔ what a base's protection asks is read from the branch, which the tools' account may read, not from the rules' list it may not (49.305635ms) [gitea] the kept token was minted with other scopes than the tools take now — minting one with them [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-FJ9Q2j/state/token [gitea] deleted 1 token(s) named "mesh-tools" the admin account "mesh-admin" held: the tools act as "mesh-tools" now ✔ the administrator's token from before goes: a kept admin token is minted again for the tools' account, and the forge drops the old one (42.872422ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-TwIXjX/state/token ✔ a tools' account someone made an administrator is made a plain account again at the next mint (36.937157ms) ✔ the tools never act as the administrator: MESH_GITEA_TOOLS_USER naming it is refused (0.8763ms) ✔ the manifest hands the tools their account, and the merge-ready authors name it rather than the administrator (1.044717ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-Cpae4b/state/token ✔ a repository check said as a warning on a head without merge-check.sh reads the base's protection from the branch (37.45392ms) [gitea] no token kept — minting one ✔ a mint that fails leaves the administrator's old token in place: the tools are never left without one (27.773352ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-zmQCEc/state/token [gitea] FAILED to delete the token(s) named "mesh-tools" the admin account "mesh-admin" holds — tried again at the next start: Gitea DELETE /users/mesh-admin/tokens/902 (named mesh-tools): 500 {"message":"the forge is not well"} ✔ an old token that cannot be deleted after the mint is said, the new token kept, and deleted at the next start (43.762187ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-h2lcrr/state/token ✔ a team another process made between the look and the create is found, not a failure (38.936211ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-p8E6kK/state/token ✔ setting the tools' account up puts main's rule right in every repository: no push, only the tools' account merges (issue 948) (37.47518ms) [gitea] no token kept — minting one [gitea] the forge already holds a token named "mesh-tools" — replacing it [gitea] minted a token for "mesh-tools" (write:repository, write:issue, read:user), kept at /tmp/gitea-tools-account-Vk5wtD/state/token ✔ a glob rule covering main keeps its statuses when main is given a rule of its own name (issue 948, review round 1) (35.260179ms) ✔ a start with the token kept still puts main's rules right: the module's start asks for it each time (issue 948, review round 1) (13.215432ms) [gitea] novox/mesh-host's main moved from a2c2c35bbbbb to d28ca95ccccc: 25 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323) [gitea] novox/mesh-host's main moved from d28ca95ccccc to ffffffffffff: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323) [gitea] novox/mesh-host: whether a trunk moved could not be weighed, so it is weighed again on the next pass — the forge is away [gitea] novox/mesh-host's main moved from a2c2c35bbbbb to d28ca95ccccc: 0 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323) [gitea] novox/mesh-controller's feat/57-config-and-provisioning moved from eeeeeeeeeeee to d28ca95ccccc: 1 open pull request(s) into it put out of date, nothing run (novox/hq ADR 0323) ✔ both statuses say the head and the trunk commit the verdict judged, first, whatever the summary's length (1.48711ms) ✔ a trunk's move puts an open head's merge check out of date, naming both commits, unless a check is on its way (0.806289ms) ✔ a push to a trunk puts every open pull request into it out of date — twenty-five of them — and none into another (2.970192ms) ✔ a trunk that cannot be read is weighed again on the next pass, never recorded half (0.287672ms) ✔ the trunk's move takes its turn after a verdict in flight, and weighs what the forge holds then (0.385335ms) ✔ a verdict judged against a trunk newer than the pass's first read is not put out of date (0.231306ms) ✔ a verdict that landed out of date keeps its account on the pull request (0.213181ms) ✔ a release branch's move puts every open pull request into it out of date, as a trunk's does (ADR 0358 §5) (0.408589ms) ✔ a signed event is published on the bus and acted on once (52.481775ms) ✔ an unsigned event, or one whose signature does not verify, is refused and published nowhere (31.605187ms) ✔ an event and the poll seeing the same head announce it once (0.305576ms) ✔ each forge event is named in the module's stream and trimmed (0.46733ms) ✔ the webhook is made a system webhook, which the forge sends every repository's events to (1.113676ms) ✔ default webhooks left at the receiver by an earlier start are replaced by one system webhook (0.990175ms) ✔ a repository's copy of a default webhook at the receiver is deleted, on every page, and said (3.078197ms) ✔ a forge that makes no system webhook is said, never taken as heard (0.675772ms) ✔ the receiver tells what it verified and what it refused (7.665953ms) ✔ a forge that never delivers is said loudly once, and unhealthy, when the poll finds what no event said (1.20583ms) ✔ a forge whose events stop after one delivery is said loudly once per silent spell, and refusals are unhealthy (0.724263ms) ✔ without a receiver by configuration the polls alone watch, healthy; a receiver that failed is unhealthy (0.343116ms) ✔ the polls look every 30 s until events are heard, every five minutes while they are, and 30 s again on a miss (0.187533ms) ℹ tests 147 ℹ suites 0 ℹ pass 147 ℹ fail 0 ℹ cancelled 0 ℹ skipped 0 ℹ todo 0 ℹ duration_ms 751.836759 ```

mesh-tools added the mesh/ready label 2026-10-16 22:14:59 +00:00

mesh-tools commented 2026-10-16 22:14:59 +00:00

Author

Member

Copy Link

Copy Source

ready at ae9cda8531

Reviewed by lead-57@novox: Phase 2 approved 6f45570 (round 3, opus); 40e40bd (ssh-client moved, same tree as ae9cda8) approved in review round 2 of issue 969; merge-gate and repo-check pass at this head on the controller with issue 969 delivered. Marked through mesh-delivery.mark-ready (novox/hq issue 913).

ready at ae9cda8531b3fbf7d7c240f15bafef8856822c4b Reviewed by lead-57@novox: Phase 2 approved 6f45570 (round 3, opus); 40e40bd (ssh-client moved, same tree as ae9cda8) approved in review round 2 of issue 969; merge-gate and repo-check pass at this head on the controller with issue 969 delivered. Marked through mesh-delivery.mark-ready (novox/hq issue 913).

mesh-tools merged commit 734a77f5a9 into main 2026-10-16 22:31:17 +00:00

mesh-tools deleted branch feat/agents-write-git-ssh 2026-10-16 22:31:17 +00:00

mesh-tools referenced this pull request 2026-10-16 23:05:18 +00:00

Let gitea's forge-account provider read the assignments it learns its consumers from (issue 955) #341

mesh-tools referenced this pull request 2026-10-16 23:26:18 +00:00

Keep gitea's forge-account provider running when a watch is refused, so a later grant is picked up (issue 955) #342

Sign in to join this conversation.

Reviewers

No Reviewers

Labels

Clear labels

mesh/ready

Reviewed and ready: mesh-delivery merges at the marked head once its checks pass (hq ADR 0364)

No labels mesh/ready

Milestone

No items

No Milestone

Projects

Clear projects

No projects

Assignees

Clear assignees

agent-ace agent-g14 agent-novox agent-shanks jschoubben mesh-tools miel.truyen nox-agent

No Assignees

1 Participants

Notifications

Subscribe

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: novox/mesh-catalog#340