skill: a census of what cannot be read is not a census of what is read wrongly

[gHashTag]

tri unparsed censuses the specs the compiler cannot read. There is no census for
the specs it reads wrongly, and this is one: 14 specs pass parse, pass
typecheck, and emit a declaration that cannot mean anything in any target language.

Minimal reproducer — four lines, and every stage accepts it

module probe {
    pub const Thing = struct {
        ok : u8,
        bad : 0,
    };
}

stage	verdict
t27c parse	ACCEPTS
t27c typecheck	ACCEPTS
gen-rust	pub bad: 0,
gen-c	0 bad;
gen-zig	field absent from the output entirely

An integer literal is in type position and no stage says a word. Two backends emit
it verbatim into languages that cannot parse it; the third drops the field silently,
which is the more dangerous of the two failures.

Where it comes from

Fourteen corpus specs use a list-valued declaration the parser does not implement:

pub const ExitCode = struct {
    kind : Enum(u8),
    variants :
      - success : 0,
      - command_error : 1,

The intent is an enum with named discriminants. What the parser builds is a struct
whose fields are kind, variants (empty type), then success with type 0,
command_error with type 1, and so on. The generated Rust is:

pub struct ExitCode {
    pub kind: Enum(u8),
    pub variants: ,
    pub success: 0,
    pub command_error: 1,

Measured

specs emitting an empty field type pub f: ,	14
of those, present in the debt ledger	3
of those, passing both parse and typecheck	14 of 14
specs with a literal in field-type position	1 (8 fields)

So 11 of the 14 are tracked by nothing at all, and all 14 satisfy every gate this
project runs.

The proposal, and its cost — your call

A typecheck rule refusing an empty field type catches all 14 and needs no
judgement: a field must have a type. But it turns 14 currently-passing specs into
typecheck failures, and docs/reports/suite_expectations.json sits at 152 entries
against a cap of 152. max_entries re-blesses at min(previous, observed), so
absorbing them needs a deliberate hand edit raising the cap — which the design treats
as exactly that, a hand edit.

The alternatives are worse or bigger:

• narrowing to "literal in type position" is unambiguous but fires on 1 spec, a rule priced at almost nothing
• having the emitters refuse instead moves generates from 581 to 567 — the same number moving down, in a more visible place
• implementing the list/enum form is a language feature, not a repair, and is not mine to author

Reporting rather than shipping. The reproducer above is the durable part.

Skill §566 and a docs/now entry. No code — the three repairs are priced in the issue and none is taken.

Refs #3225

[github-actions]

📓 NotebookLM Notebook linked to this PR

• Notebook ID: b83263109fb055dc
• URL: https://notebooklm.google.com/notebook/b83263109fb055dc

This notebook contains session context, decisions, and artifacts for this work.

[github-actions]

PR Dashboard

Generated at: 2026-09-05 01:52:47 UTC

• prcheck: a completeness guard asked of the filtered half — it could only ever say COMPLETE #3224 prcheck: a completeness guard asked of the filtered half — it could only ever say COMPLETE | Created: 2026-09-05 | Updated: 2026-09-05 | 3 failing / 30 total
• fix(rust): lower Zig's undefined to Default::default() (+2) — NOT auto-merged, needs your call #3223 fix(rust): lower Zig's undefined to Default::default() (+2) — NOT auto-merged, needs your call | Created: 2026-09-05 | Updated: 2026-09-05 | 6 failing / 36 total
• gates mutate: X of Y where X can exceed Y — four lines from a fix I shipped two passes ago #3221 gates mutate: X of Y where X can exceed Y — four lines from a fix I shipped two passes ago | Created: 2026-09-05 | Updated: 2026-09-05 | 5 failing / 35 total
• skill lost: a body that is not its own — and the prototype that was right because one alternative was dead #3214 skill lost: a body that is not its own — and the prototype that was right because one alternative was dead | Created: 2026-09-05 | Updated: 2026-09-05 | 1 failing / 36 total
• docs: the format catalog count is a live invariant, not a fixed number #3186 docs: the format catalog count is a live invariant, not a fixed number | Created: 2026-09-04 | Updated: 2026-09-04 | 4 failing / 27 total
• chore(ci)(deps): bump actions/attest-build-provenance from 1 to 4 #3001 chore(ci)(deps): bump actions/attest-build-provenance from 1 to 4 | Created: 2026-09-01 | Updated: 2026-09-05 | 1 failing / 29 total
• chore(ci)(deps): bump softprops/action-gh-release from 2 to 3 #3000 chore(ci)(deps): bump softprops/action-gh-release from 2 to 3 | Created: 2026-09-01 | Updated: 2026-09-05 | 1 failing / 29 total
• chore(ci)(deps): bump github/codeql-action from 3 to 4 #2999 chore(ci)(deps): bump github/codeql-action from 3 to 4 | Created: 2026-09-01 | Updated: 2026-09-05 | 1 failing / 28 total
• Land the formal verification harness on master (it has never run) #2727 Land the formal verification harness on master (it has never run) | Created: 2026-08-27 | Updated: 2026-08-27 | 5 failing / 24 total
• chore(deps)(deps): bump thiserror from 2.0.18 to 2.0.20 #2693 chore(deps)(deps): bump thiserror from 2.0.18 to 2.0.20 | Created: 2026-08-24 | Updated: 2026-08-28 | 3 failing / 34 total
• chore(deps)(deps): bump hyper from 1.10.1 to 1.11.0 #2692 chore(deps)(deps): bump hyper from 1.10.1 to 1.11.0 | Created: 2026-08-24 | Updated: 2026-08-28 | 3 failing / 33 total
• chore(deps)(deps): bump base64 from 0.22.1 to 0.23.1 #2691 chore(deps)(deps): bump base64 from 0.22.1 to 0.23.1 | Created: 2026-08-24 | Updated: 2026-08-28 | 3 failing / 34 total
• chore(deps)(deps): bump serde from 1.0.228 to 1.0.229 #2690 chore(deps)(deps): bump serde from 1.0.228 to 1.0.229 | Created: 2026-08-24 | Updated: 2026-08-28 | 3 failing / 34 total
• chore(deps)(deps): bump uuid from 1.23.2 to 1.25.0 #2689 chore(deps)(deps): bump uuid from 1.23.2 to 1.25.0 | Created: 2026-08-24 | Updated: 2026-08-28 | 3 failing / 34 total
• Unbreak FPGA CI and correct two unbacked hardware claims #2081 Unbreak FPGA CI and correct two unbacked hardware claims | Created: 2026-08-11 | Updated: 2026-08-20 | 5 failing / 27 total

Summary

Status	Count
Total Open PRs	15
PRs with Failing Checks	11
PRs with All Checks Green	4
READY	0
FAILING	11
PENDING	0

Seal Status

• ⚠️ STALE -- sha256(compiler.rs)=9f4e7ae386ba != 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).