kyle/boxmaker
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs: the M5c-1 device check on straylight, rift and the owner's Mattermost

Browse Source 
Every M5c-1 item of the spec's section 13 passed against a build of m5c1 with the review's fixes:
clean, untrusted and secret turns routed as decided, a remembered fact concluded at once and
deleted by undo, a secret session's memory write skipped, and an outage waited out and drained
with nothing sent twice. The check's three rift workspaces were deleted afterwards.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
This commit is contained in:
kyle
2026-09-28 15:39:57 -07:00
co-authored by Claude Opus 5.5
parent 572b582038
commit ab36e54273
1 changed files with 21 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/implementer-log.md
+21 -3
View File
@@ -881,12 +881,11 @@ proved a failing `brokerd` test flaky before re-running the gate instead of chan
Accepted, with fixes made in review (below). Tasks 01 to 15 by Ornith-1.5-35B-A3B through Accepted, with fixes made in review (below). Tasks 01 to 15 by Ornith-1.5-35B-A3B through
`tools/run-plan.sh`, run on talos with no intervention, 2026-09-28 10:11 to 13:42 (about 3.5 `tools/run-plan.sh`, run on talos with no intervention, 2026-09-28 10:11 to 13:42 (about 3.5
hours). The branch differs from the reference (`~/boxmaker-refs/m5c1-ref.bundle` on straylight, hours), device-checked on straylight the same day. The branch differs from the reference (`~/boxmaker-refs/m5c1-ref.bundle` on straylight,
`c1345e4`) in six files of `memoryd::honcho` only (`client.rs`, `config.rs`, `ledger.rs`, `c1345e4`) in six files of `memoryd::honcho` only (`client.rs`, `config.rs`, `ledger.rs`,
`mirror.rs`, `report.rs`, `worker.rs`), each difference an equivalent rewrite; `config.rs`'s `mirror.rs`, `report.rs`, `worker.rs`), each difference an equivalent rewrite; `config.rs`'s
`check_url` and `check_keys` were checked by hand on an empty port, a second colon and a key that `check_url` and `check_keys` were checked by hand on an empty port, a second colon and a key that
is neither an own nor a read key. The device check (spec, section 13, the M5c-1 part) is still to is neither an own nor a read key.
do.
| Check | Result | | Check | Result |
|---|---| |---|---|
@@ -908,6 +907,25 @@ do.
| 8 | low | Ornith (task 12) | A "before" block did not match after its own earlier edit reordered imports; it anchored on its own line and logged no deviation. | Noted; AGENTS.md rule; tip T54 | | 8 | low | Ornith (task 12) | A "before" block did not match after its own earlier edit reordered imports; it anchored on its own line and logged no deviation. | Noted; AGENTS.md rule; tip T54 |
| 9 | nit | plan (every task) | Every task said to run `git switch m5c1`, which AGENTS.md forbids and the harness refused. The gate count in every task's stop list did not say it includes the ignored `end_to_end` test, which cost tasks 03, 13 and 15 much of their time. | Tips T50, T51; AGENTS.md now defines the count | | 9 | nit | plan (every task) | Every task said to run `git switch m5c1`, which AGENTS.md forbids and the harness refused. The gate count in every task's stop list did not say it includes the ignored `end_to_end` test, which cost tasks 03, 13 and 15 much of their time. | Tips T50, T51; AGENTS.md now defines the count |
Device check, 2026-09-28, on straylight (a `nix build` of `m5c1` at `572b582`, the review's fixes
included; Ornith slot 0 on straylight; the owner's Mattermost; rift, in three workspaces made for
the check, `boxmaker-check-trusted`, `-untrusted` and `-fleet`, deleted afterwards; no other
workspace touched). Every item of the spec's section 13 for M5c-1 passed:
| Check | Result |
|---|---|
| Start with `[honcho]` | the ledger's first line is `began`; `bxctl memory sync` shows `0 decided` |
| A clean turn | decided trusted, speaker `kyle`, sent in 140 ms: two messages in `boxmaker-check-trusted`, from `kyle` and `boxmaker-check`, with the turn and session in their metadata |
| "Remember that my favourite colour is teal" | `add user.md`; a conclusion about `kyle` in `<session>_memory`, sent 135 ms after the write, while its turn still ran (review finding 3) |
| `bxctl memory undo` of it | `unconclude` decided and the conclusion deleted within 20 ms; the workspace has no conclusion |
| A turn that read a page under an untrusted grant | its tool result labelled untrusted; the turn in `boxmaker-check-untrusted` only |
| A session that read a file graded `secret`, then "remember that the Aldridge renewal is on 14 March" | both turns `skip`, "the session is secret"; the write (a note journalled with class `secret`) `skip`, "the write is from a secret session" (review finding 1); nothing from the session in any workspace |
| Honcho unreachable (`memoryd` restarted on a closed port), then a clean turn | `1 waiting`; sync shows `honcho unreachable since …`; one log line with its pointer over the retries |
| Back on rift | the waiting turn sent once; sync `9 decided: 6 sent, 0 waiting, 0 failed, 3 skipped`; the first session still holds exactly its four messages after two restarts |
Not checked on the device: a refused key (rift has auth off; unit-tested) and a fleet note (no
untrusted session wrote one; unit-tested).
What was good: every given file and every whole edit arrived exactly as written, fifteen tasks in What was good: every given file and every whole edit arrived exactly as written, fifteen tasks in
a row with no stop, including task 15's thirty edits over seventeen test files. Where Ornith filled a row with no stop, including task 15's thirty edits over seventeen test files. Where Ornith filled
bodies from comments, the code agreed with the reference in behaviour everywhere checked. All six bodies from comments, the code agreed with the reference in behaviour everywhere checked. All six