Skip to content

fix: repo-relative t27a seals + parseable orbitofrontal_value port (master red gates) - #6861

Merged
gHashTag merged 2 commits into
masterfrom
fix/master-red-seals-orbitofrontal
Oct 6, 2026
Merged

gHashTag merged 2 commits into
masterfrom
fix/master-red-seals-orbitofrontal

Conversation

@gHashTag

@gHashTag gHashTag commented Oct 6, 2026

Copy link
Copy Markdown
Owner

Cause

Master is red on the t27c Railway lab in three gates (suite, specs-generate, seal-coverage) since two merges:

  1. feat(isa): t27a MVP -- TRI-27 assembler/disassembler round-trip (Refs #6507) #6789 committed three seals whose spec_path is an absolute local worktree path (/Users/.../worktrees/agent-.../specs/...) instead of a repo-relative one:

    • .trinity/seals/isa_T27a.json
    • .trinity/seals/isa_Tri27Encoding.json
    • .trinity/seals/theory_CompilerTheoryIsaRoundTrip.json

    On the lab that path does not exist, so seal-coverage reports "3 seal(s) newly do not hold".

  2. Port gHashTag/trinity:src/brain/orbitofrontal_value.zig (Zig, 2 functions) to specs/port/trinity/src/brain/orbitofrontal #6755 added specs/port/trinity/src/brain/orbitofrontal_value.t27, which does not parse: a switch with block arms (.gf16 => { ... },) fails with Unexpected token in expression: LBrace ('{') at line 96:18. That one spec fails specs-generate (gen_hash=none) and the suite ratchet (1 unexpected [parse] failure).

Fix

  • The three seals now carry repo-relative spec_path like every other seal. Every hash in a seal is computed from the spec source only (compute_seal_hashes never hashes the path), and each spec_hash was re-checked against the current spec file, so this is a path-only change.
  • orbitofrontal_value.t27: the four switch arms are rewritten as an if / else-if chain over the enum tag (recommended == .gf16, ...), same values per arm; PHI_INV is hoisted out of the arm. Tests unchanged. The parser is not touched.

Only .t27 and tool-written seal data change.

Refs #6507
Refs #6488

🤖 Generated with Claude Code

- .trinity/seals/{isa_T27a,isa_Tri27Encoding,theory_CompilerTheoryIsaRoundTrip}.json
  (from #6789) stored spec_path as an absolute local worktree path, so
  seal-coverage could not find the specs on the lab. spec_path is now
  repo-relative like every other seal. All seal hashes are content-only
  (spec_hash re-checked against the current specs), so only the path
  changes.
- specs/port/trinity/src/brain/orbitofrontal_value.t27 (from #6755) did not
  parse: a `switch` with `.tag => { ... }` block arms hits "Unexpected token
  in expression: LBrace". The four arms are rewritten as an if / else-if
  chain over the enum tag with the same values; tests unchanged.

Refs #6507
Refs #6488

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@github-actions

github-actions Bot commented Oct 6, 2026

Copy link
Copy Markdown
Contributor

PR Dashboard

Generated at: 2026-10-06 14:03:07 UTC

Summary

Status Count
Total Open PRs 50
PRs with Failing Checks 41
PRs with All Checks Green 9
READY 0
FAILING 41
PENDING 0
NO CHECKS YET 1

These columns do not partition: 0 + 41 + 0 + 1 = 42, and there are 50 open PRs. A PR is being counted twice or not at all.

Seal Status

  • ⚠️ STALE -- sha256(compiler.rs)=3c78f3c7ffb7 != manifest seal=87e5cbd3ad94.
    The committed NMSE numbers were certified against an older compiler.rs.
    Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

This was referenced Oct 6, 2026
Master's #6862 already resealed the three t27a seals with repo-relative
paths; take master's copies. Only the orbitofrontal_value.t27 parse fix
remains in this PR.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@github-actions

github-actions Bot commented Oct 6, 2026

Copy link
Copy Markdown
Contributor

PR Dashboard

Generated at: 2026-10-06 14:57:57 UTC

Summary

Status Count
Total Open PRs 50
PRs with Failing Checks 36
PRs with All Checks Green 14
READY 0
FAILING 36
PENDING 0
NO CHECKS YET 8

These columns do not partition: 0 + 36 + 0 + 8 = 44, and there are 50 open PRs. A PR is being counted twice or not at all.

Seal Status

  • ⚠️ STALE -- sha256(compiler.rs)=0ec05a8c1a46 != manifest seal=87e5cbd3ad94.
    The committed NMSE numbers were certified against an older compiler.rs.
    Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

@gHashTag
gHashTag merged commit cf84846 into master Oct 6, 2026
24 of 27 checks passed
gHashTag added a commit that referenced this pull request Oct 6, 2026
The branch sat on #6894's base, which predates #6861 (parseable
orbitofrontal_value port); the lab's specs-generate and suite ratchet were
red on that inherited spec only. Conflict in foreign-exceptions.txt: both
sides kept. Published figures re-pinned to the merged corpus: test blocks
15697 (master 15684 at 31ad9df + #6894 8 + #6940 5), x.len 2149
(master +2).

Refs #6940

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant