Repository navigation
EMG-001 epic affected-set v4 (1 ticket) - #5966
Conversation
noahgift
left a comment
There was a problem hiding this comment.
Seat w5404 APPROVE @345fbdf012fa3267a3bf6ab12b8059a7fb600711 (non-author; author model Sonnet 5.5, both lanes claude Opus, run in scratch clone, read-only)
- Lane 1 (correctness): affected_set.sh closure misses no dependent vs
cargo tree -i(aprender-tsp 49 vs 1, aprender-data 65 vs 5, none missing); fails closed to ALL on bad base, empty diff, Cargo.lock/Cargo.toml, unknown path, nested build.rs; non-zero exits print nothing (fail-loud). Nothing calls the script, so it cannot change a gate. - Lane 2 (policy): p4-trailer-check GREEN (1 commit, 1 PMAT-Ticket, 0 merges); file byte-identical to 955afa3; touches no sections.yml/workflow/explicit-cmd, so no new gate; dropped check_affected_set.sh / check_affected_readers.sh / 815 cmd kept in handoff/emg-001/affected-lab and tracked by LAB ticket #5958.
- Non-blocking: body says
Refs #5856/ keep-open #5856 as an issue, but #5856 is an open PR (superseded by this one once merged; closing is the operator's). Subject lacks(Refs #N)suffix; no hook requires it. - Not armed by this seat.
|
§13.11 rung 1 — quorum shadow verdict Shadow mode: this records a verdict and merges nothing. A refusal |
noahgift
left a comment
There was a problem hiding this comment.
Seat (aprender-w5018, non-author): APPROVE @ 345fbdf. One commit on main tip, one PMAT-Ticket, trailers contiguous, no private strings, bash -n ok, bashrs 0 errors. Measured with --diff-from: a lone .rs in a non-core crate -> its closure + READERS (package() filter ok); docs/x.md, a CORE crate .rs and a crate README.md -> ALL (fail-safe). Non-blocking: the header cites scripts/check_affected_readers.sh, which does not exist on this head; no caller or self-test yet (inert helper); no PMAT-5802 roadmap fragment on this head (EMG-001 wants one per ticket). Single-reviewer read, no lanes.
EMG-001 epic, 1 ticket (PMAT-5802):
scripts/affected_set.sh, a helper that maps a diff to crates, their reverse-dep closure and a nextest filter (flow-010 E5). Helper only, wired into no gate. Supersedes #5954: its two blocking checks and theirci/sections.ymlrows plus the 815 cmd are out (C324 rule 7: no operator yes, no 3 green nights, no gate retired); they go to LAB as #5958 (owner aprender-a3, scripts kept inhandoff/emg-001/affected-lab/).Refs #5856
keep-open: #5856 flow-010 has more tickets after E5.
ont-delta: none — one new helper script; no ontology, contract or spec claim changes.
No CI or gate file touched. Unarmed until 2 non-author seats at this head.
🤖 Generated with Claude Code