provenance sweep sub-topic: repo-wide sweep, upstream-drift convention engagement, and the cleaned-state gate

The provenance build plan's Phase 8 promotes the repo-wide sweep, the combined upstream-drift
convention engagement, and the narrow CI regression gate for the cleaned state into their own
sub-topic. This issue is that promotion artifact, and it is the Brief the sub-topic runs from.

Why this is an issue rather than `docs/topics/provenance-sweep/PLAN.md`

The plan named a contract slice as the promotion target. That target is not merge-able:
docs/topics/<slug>/ is Contract tier, branch-only, and scripts/check-contract-slice-prune.sh
hard-fails any new path there on the way to main — the parent topic's own slice was pruned for
exactly this reason (#3456). The work-item tracker is the graduation route the topic-docs
lifecycle names for actionable follow-ups, so the Brief lands here. Nothing about the plan's
content changes; only its vessel.

Precondition, and it is a hard one

The sweep cannot apply dispositions yet, and this is the gate's own arithmetic rather than a
blocker to route around. The golden set ships at 10 cases across four groupings, so no class
reaches min_n_per_class 10 and every class is report-only at v1. fix and sweep apply
dispositions to fix-eligible findings only, and at v1 there are none.

So the sub-topic has two distinct stages, and the first does not wait on the second:

A report-only sweep is runnable now and is worth running: it produces the corpus-scale
telemetry (fetch counts, cache hits, carve-out declines, not-found rates) that the convention
engagement needs as evidence, and it harvests adjudications into golden cases.
A disposition-applying sweep unlocks only once the growth loop carries verbatim and
near-verbatim past n=10 at or above the 0.95 precision bar. That is the named exit condition
of the first growth round in provenance plugin: phases 6 to 8 (golden set and measurement, crosswalk registration, sweep) #3458.

Running stage 1 and feeding its adjudications back is what makes stage 2 reachable. Reversing
them is not an option.

Measured baseline entering the sweep

From the Phase 6 measurement (recorded in plugins/provenance/CHANGELOG.md): over 1,347 tracked
markdown files after carve-outs, 525 stamp candidates, 482 parsed, 43 declined, 0 expired at the
180-day default. Golden set: 10 cases, overall 8 tp / 0 fp / 0 fn / 2 tn, every class report-only.

The adversarial rotation probe (design thread T15) returned a real limit worth carrying into the
sweep: at roughly one substitution every four words, containment goes to 0.0 with no matched
spans against a source that was fetched and identity-checked. Word-shingling is evadable by an
author who intends to evade it, and no value of min_containment above zero recovers a passage
with zero matching shingles. The sweep should expect a floor on what the deterministic half can
confirm, and should not read a clean fingerprint as proof of originality.

Execution contract (inherited from the parent Brief)

One tracked file at a time. Apply the verdict, verify (semantic diff plus pointer liveness),
close.
A file is closed when every finding in it carries a disposition or an explicit neutral
outcome — never when the interesting ones are done.
Record each closure in the sweep ledger so an interrupted sweep resumes without re-deciding
closed files, and so the closure count is a fact rather than a memory.
Carve-outs stay categorical. If the sweep starts accumulating per-instance suppressions, that is
evidence a carve-out is drawn wrongly, and the fix is to redraw it in rubric.md — not to keep
a suppression list. This matters beyond tidiness: see the convention engagement below.

The two additions the parent plan requires this Brief to carry

The live no-breadcrumb (not-found) probe, as a first-run validation item. That disposition
has never been exercised live — every spike source was supplied by hand. The sweep's first
genuine no-breadcrumb candidate validates that the searched-surfaces listing names every
surface actually checked, and that budget exhaustion reads as a first-class neutral outcome
rather than as an acquittal.
Sweep resume semantics. The fetch ceiling and the response cache are scoped per SWEEP, not
per invocation, and both are recorded in the closure ledger. A resumed sweep therefore neither
resets its spend nor silently reuses a stale cache. The ledger is checkout-local; state that
explicitly, because a sweep resumed in a different checkout is a new sweep.

The convention engagement (fires at sweep completion only)

Drafted in full at docs/specs/provenance-convention-engagement.md. It executes as one entry
in docs/conventions/upstream-drift/CHANGELOG.md, carrying three parts:

Recorded-decision re-derivation. chore(conventions): plugins restating upstream harness facts are not adopters of upstream-drift, so stale claims never surface #2273 deferred an adoption gate with its own stated
trigger: "a detector is demonstrated that separates an upstream restatement from an in-repo one
without a suppression list." Sweep completion either is that demonstration or it is not, and
the engagement says which from evidence — the closure ledger, the measured precision/recall at
the stated minimum n, and the carve-out record showing categorical exclusions only. If the
sweep accumulated per-instance suppressions, the bar is not cleared, the trigger has not fired,
and the engagement records exactly that with the count.
The honest expected outcome, stated now so the engagement cannot overclaim later: the detector
is LLM-led with deterministic verification, so the enforceability row moves at most from
reasoning-only to detect-then-judge, and a deterministic suppression-free CI gate in the
*-gate pattern remains unavailable.
Enforceability table update, version bump conditional. Major bump ONLY if an enforceability
verdict changes; otherwise a minor entry with the as-of date refreshed and the outcome stated,
drift or no drift.
The trigger-less-stamp table-row flip. Partially pre-consumed: the check's code shipped in
Phase 3 and its crosswalk row in Phase 7 (feat(provenance): documentation-provenance audit plugin (phases 1 to 5) #3456 and the Phase 7 work). What remains here is
exactly the convention's own row moving from "named but not built" to "built, off-by-default,
repo-override enabled", plus the build-trigger bookkeeping. The registry row's early presence
is scheduled, not drifted.

Also recorded at the same firing: the hash-store designed-issue trigger evaluated against the
sweep's fetch telemetry; the dead-pointer round-trip wiring (the weekly link-check lane routing
to the demotion path in reference/dispositions.md); and the adopters table, where rows are
added only for surfaces that actually conform, with #2297 getting its closure evidence if the
repository ends the sweep conforming.

What the engagement never does: declare the migration playbook's plugin-eval deferral trigger
fired; rewrite prior changelog entries or dated records; or convert a judgment verdict into a
deterministic claim — report-only tiers stay report-only whatever the sweep measured.

Checklist

Grow the golden set past min_n_per_class for verbatim and near-verbatim (provenance plugin: phases 6 to 8 (golden set and measurement, crosswalk registration, sweep) #3458)
Run the report-only sweep; record closure ledger, telemetry, and adjudications
Validate the live no-breadcrumb not-found probe on its first genuine case
Confirm sweep resume semantics (per-sweep ceiling and cache, checkout-local ledger)
Disposition-applying sweep, once a class is fix-eligible
Execute the convention engagement as one upstream-drift/CHANGELOG.md entry
Evaluate the hash-store designed-issue trigger against the sweep's fetch telemetry
Wire the dead-pointer round-trip to the weekly link-check lane
Narrow CI regression gate for the cleaned state

provenance plugin: phases 6 to 8 (golden set and measurement, crosswalk registration, sweep) #3458 — phases 6 and 7, and the golden-set growth this sub-topic depends on
feat(provenance): documentation-provenance audit plugin (phases 1 to 5) #3456 — phases 1 to 5, where the plugin, its scripts, and the trigger-less check shipped
Refs chore(conventions): plugins restating upstream harness facts are not adopters of upstream-drift, so stale claims never surface #2273 — the upstream-drift recorded decision this engagement re-derives
Refs chore(plugins): 12 unstamped upstream harness claims across 11 plugins — upstream-drift adoption backlog #2297 — the fleet's open-carriers issue, which gets its closure evidence here
docs/specs/provenance-convention-engagement.md — the engagement, drafted in full
docs/specs/provenance-design-threads.md — T15, the adversarial probe whose limit is recorded above

feat(provenance): golden set, crosswalk registration, and two defects…

fix(provenance): correct a stale Phase 6 baseline, and pin its as-of …

feat(provenance)!: rubric v3 states the scope C3 and C4 are graded at

Stage 1 report-only sweep: first run, recorded

Ran the report-only sweep across the whole corpus at d7e391da, split into two judgment legs (docs/, plugins/) plus a mechanical telemetry leg, each with an independent fresh-context verifier. No dispositions were applied and no tracked file was edited by any sweep leg: at v1 no class reaches min_n_per_class, so applying one would be a correctness violation rather than a shortcut.

Checklist movement

Run the report-only sweep; record closure ledger, telemetry, and adjudications — below.
Validate the live no-breadcrumb not-found probe on its first genuine case — validated, with a gap found.
Confirm sweep resume semantics — measured, and the answer is that they are not implemented.
Grow the golden set past min_n_per_class — candidates harvested below; not yet converted.
Disposition-applying sweep — still blocked, and now further away (see rubric v3).
Convention engagement, hash-store trigger, dead-pointer wiring, cleaned-state gate.

Telemetry

metric	value at `d7e391da`, `--as-of 2026-08-28`
tracked markdown after carve-outs	1,352 (1,395 considered, 43 declined at path level)
stamp candidates / parsed / declined	535 / 491 / 44 (20 month-name, 24 bare-year)
expired at 180-day default	0
oldest parsed stamp	2026-04-08, `plugins/work-items/skills/track/actions/add.md:104`
findings at a 60-day window	9
fetches attempted / succeeded / failed	9 / 7 / 2 (both HTTP 404)
rate-limit failures (429)	0
fingerprint comparisons	784
categorical carve-outs in effect	3
per-instance suppressions accumulated	0

Two of these figures are date-relative and expire, which is why the as-of date is pinned: the 180-day zero holds only until 2026-10-05 on the current oldest stamp, and the 60-day count moves daily.

The adoption-bar evidence this issue asks for

This issue states that if the sweep accumulates per-instance suppressions, the #2273 adoption bar is not cleared. It accumulated none. Three categorical carve-outs, no per-instance suppression list anywhere, and two skill surfaces forbid them by design. That is the first evidence on record pointing at the trigger, and it is recorded here rather than in the convention, because the engagement fires at sweep completion and this is one run.

The excluded_paths placement holds as designed: the fixtures exclusion lives in the config layer, not unconditionally in list-corpus.sh, so the eval harness lifts it while a normal run declines it and says so. The #3041 precedent is live.

Resume semantics are not implemented

All three claims in this Brief measured as NOT IMPLEMENTED:

claim	verdict	evidence
fetch ceiling scoped per sweep	not implemented	`reference/source-fetch.md:110` scopes `corpus_fetch_ceiling` to "the run"
response cache scoped per sweep	not implemented	`source-fetch.md:113-115`, cache "lives in the run's memory slice"
both recorded in the closure ledger	not implemented	`reference/dispositions.md:152-154`, `SKILL.md:163-165` record closures and guard outcomes only

Checkout-local holds by design (docs/specs/provenance-plugin-topology.md:60 places the ledger under .work/, which .gitignore:30 ignores) but no script enforces it — .work/ is referenced by no script at all. The Brief describes intended semantics; whether they were built is what this checklist item measured, and they were not. Recording rather than building, since this run was scoped to measure.

Findings

docs/ (148 files) does not lexically copy the upstream it cites. Exhaustive 296-comparison sweep, corpus-wide max containment 0.011. The five docs/upstream/*.md records tested against a real clone of mattpocock/skills scored 0.001–0.009 with zero firings, and their self-declared near-verbatim adoption cites upstream to the correct line.

plugins/ produced six findings that stand, four of them in one file:

location	class	tier
`plugins/context7/skills/lookup/context/lookup.md:40-41`	verbatim	fingerprint-confirmed
same file, three further spans	near-verbatim	see caveat below
`plugins/firecrawl/skills/firecrawl/context/commands.md:36`	paraphrase	llm-suspected
`plugins/playwright/skills/playwright/reference/test-generation.md:80`	near-verbatim	source-fetched-similar

lookup.md reproduces a five-step procedure from the live upstream in order, containment 0.217 with a 26-word span, and carries zero attribution and zero external URLs across 77 non-empty lines. The plugin does attribute upstash — in a sibling file — and C3 grades what is on the page.

Tier caveat, found by the verifier: two of those four were tiered on the file-level separation flag, but their per-finding longest spans are 14 words each, below the 15 floor, and file containment 0.217 is below 0.3. Neither arm fires for them individually. They are genuine short copies below the deliberate floor — the floor working as designed — and must not be recorded as fingerprint-confirmed. Related process deviation: the cited ranges match no module span, so nomination ranges were carried through unrefined against SKILL.md step 7.

A materialized drift, outside the tier table's domain: docs/NATIVE-SURFACES.md:98 records a native description absent from both shipped claude binaries in this container. The row's own recheck trigger has already fired. This is not drift risk; it is drift. check-stamps.sh structurally cannot see it.

T15 reproduced live, twice

The rotation limit is no longer only a synthetic probe. firecrawl/.../commands.md:36 restates (default 5, max 100) from the Firecrawl CLI docs at containment 0.000 with zero matched spans, against a source that was fetched and identity-checked. The verifier control-tested the module rather than trusting the zero (local-vs-itself returns 0.927), so this is a genuine detector miss on a table-row-to-prose restatement, not an operational failure. llm-suspected is the forced tier: no lexical evidence is possible.

Neither leg proposed moving the 0.3/15 constants, and neither read a clean fingerprint as proof of originality.

The no-breadcrumb probe validated, and found a gap

Resolved at rung 4 (local binaries) after a zero-URL file. Budget exhaustion was exercised on a second candidate (404 then 403) and held as a neutral outcome, not an acquittal — correctly, because a further rung then resolved it, and stopping early would have lost the materialized drift above as a false negative.

The gap: the searched-surfaces listing has no mechanical floor. emit-findings.sh has no field for it, no budget block, no schema. Three prose requirements, zero enforcement — a listing omitting a rung is indistinguishable downstream from a complete one.

Also unplanned and worth a decision before stage 2: the stale string's authoritative home is docs/native-surfaces/records.json:108, outside the markdown corpus and generated, and its rendering says never to hand-edit it. A fix pass would edit a generated file.

Detector defects found by running it on itself

False negative, being fixed: docs/upstream/aihero-course.md:127 declines as "bare year" while ending in a conforming ISO date — a real stamp silently exempted from expiry.
False positive, found independently by both legs: the ordinary English word "may" read as a month name (docs/conventions/standards/README.md:134, plugins/claude-config/skills/audit/SKILL.md:171). Also SC2034 and "cache_read_input_tokens": 2000 read as bare years.
The declined block's text field reports the keyword line rather than the date line, so whoever adjudicates the 44 declines sees misleading context.

Adjudication candidates harvested

Rejected findings are the high-value output, because a case converted from a rejected finding is one nobody authored to pass. Harvested, not yet converted: a byte-identical 16-word list of eight product-defined identifiers that fires the separation rule but fails C2 (no other way to write it); a conforming record whose trigger has fired, a shape no golden case has; carve-out 5's reach over RESEARCH-*.md; and a 0.589-containment/142-word case that turned entirely on grading scope.

That last one produced a rubric change: C3 and C4 are graded at different scopes and v2 never said so. Two readers agreed on the verdict and disagreed on which scope produced it. Rubric v3 states it — C3 outward across the file, C4 on the passage — and takes the invalidation: the v2 golden-set measurement does not carry forward, so the set must be re-scored before any precision figure is cited. That pushes fix-eligibility further out, not closer.

Honest limits on this run

Files were enumerated, not model-read. Model-led nomination covered roughly 35 files in plugins/; the rest got path- and stamp-level treatment only.
nomination_passes ran once, not twice — nominations were not unioned.
judge_samples was not met: judgments are single-context, not three blind panels. No verdict here carries panel unanimity. Both judgment legs reported this themselves rather than being caught at it.
Recall is therefore lower than designed. Moot for eligibility, since no class reaches minimum n either way, but it means the absence of findings in a tree is weaker evidence than the presence of one.
One leg's verifier caught a wrong-artifact identity error (two claude binaries in this container, 2.1.246 and 2.1.250; the finding survived at both, but the pin was wrong). Another leg could not obtain a verifier at all and said so rather than claiming one.

Recommended before stage 2

Decide the generated-file question — a fix pass must not edit records.json through its rendering.
Give the searched-surfaces listing a mechanical floor, or stop claiming it as a validation.
Re-score the golden set against rubric v3.
Add a tier row for a vendored-snapshot source: strong provenance, weak currency, and every current row gates on "fetched" or "no source".

Generated by Claude Code

fix(provenance): supply the judge the file rubric v3 tells it to grade

fix(provenance): apply the window rule and the v3 file input to both …

docs(provenance): correct three counts this entry got wrong about itself

fix(provenance): four defects the first report-only sweep found, thre…

provenance/audit: vendored-snapshot tier rule; record searched-surfac…

fix(provenance): stop the rubric from carrying its own answer key

feat(provenance): schema-check the not-found searched listing, and st…

fix(provenance): remove two answer-key leaks from the golden set, clo…

This was generated by AI during triage.

Claiming for triage-only evaluation (no execution).

Generated by Claude Code

This was generated by AI during triage.

Triage: verified, oversized, routed human-gated with a decomposition proposal

Stage 1 ran and is recorded in detail. What this pass adds is a current-state re-measurement, which moves two things, and an assessment that the item is too large to route as one unit but cannot yet be sliced cleanly.

Re-measured on current `main`, `--as-of 2026-09-06`

files 1,471 | candidates 629 | parsed 585 | declined 44 | findings 87
expiry_days 180 | trigger_less_check TRUE

The trigger-less check is on and producing 87 findings. All 87 carry rule-trigger-less-stamp; zero are expiry findings, so the sweep's "0 expired at the 180-day default" still holds. This is directly load-bearing for the checklist's third convention part: the row's move to "built, off-by-default, repo-override enabled" is not pending, the override is live, and 87 is the number the engagement will have to report. Worth capturing before the figure moves.

Two of the recorded detector false positives are still present. Reproduced from the declined groups:

location	declined as	actual content
`plugins/code-tidying/.../shell-tooling.md:38`	bare year	`SC2034`, a shellcheck code
`plugins/context-guard/reference/reader-contract.md:78`	bare year	`"cache_read_input_tokens": 2000`

A third is questionable: a table row in docs/upstream/mattpocock-skills-v12-map.md:48 declines as a bare year with no year in view. And one decline carries a wrong reason string: plugins/github/reference/recipes/billing.md:140 reads As of 2026-07 and is declined as "bare year, no month or day" when it plainly has a month.

Two recorded defects appear resolved. The ordinary-English "may" false positive no longer appears in any month-name decline (all five are genuine month-name forms, three of them the plugin's own changelog describing this very bug). The text-field defect is not visible either: every example now shows the matched line. The stamp under-report fixed in #3473 is the likely cause. The issue's checklist does not track these, so their resolution is recorded here rather than ticked.

Net: the decline classifier still has a real precision problem across 39 bare-year declines, narrower than at sweep time but not closed.

Why this is oversized

Nine checklist items, three complete. The remainder are not variations on one task, they are five different kinds of work:

Detector defects — the surviving bare-year false positives and the wrong reason string. Small, testable, no open question.
Unbuilt mechanism — resume semantics measured as NOT IMPLEMENTED on all three claims, and the searched-surfaces listing having no mechanical floor. The Brief describes intended behavior that was never built, so this is feature work, not repair.
Open spec decisions — four of them, listed below.
A blocked stage 2 — gated on golden-set growth past min_n_per_class (provenance plugin: phases 6 to 8 (golden set and measurement, crosswalk registration, sweep) #3458) and on re-scoring against rubric v3, whose scope change invalidated the v2 measurement. Two preconditions, not one, and the second was added after this issue was written.
A convention changelog entry plus a CI gate — both fire only at sweep completion.

Why it is not being decomposed in this pass

Decomposition needs stable slice boundaries, and four unanswered decisions from the stage-1 run sit directly on them:

The generated-file question. The materialized drift's authoritative home is outside the markdown corpus, is generated, and its rendering says never to hand-edit. A fix pass would edit a generated file. Until this is answered, nobody knows whether the sweep's disposition scope includes generated surfaces, which changes what a stage-2 slice even contains.
The searched-surfaces listing. Three prose requirements, zero enforcement. Either give it a mechanical floor or stop claiming it as a validation. The checklist's no-breadcrumb item is marked complete against a validation that is unenforced, so this decides whether that tick stands.
Rubric v3 re-scoring scope. The invalidation is recorded but its extent is not: whether all ten cases re-score or only the C3/C4-sensitive ones determines whether this is a slice or a footnote.
The vendored-snapshot tier row. Every current row gates on "fetched" or "no source". Adding a row changes the tier table, which the golden set is scored against, which feeds min_n. Sequencing this after the re-score wastes the re-score.

Slicing before these are answered produces items that get rewritten as soon as they are. That is the specific failure decomposition is supposed to prevent, so this pass stops at the proposal.

Proposed decomposition, for approval in one pass

Blockers first. A maintainer answering the four decisions above can approve this split as-is.

#	Slice	Depends on	Shape
A	Fix the surviving bare-year false positives and the wrong reason string	none	agent-ready, scoped
B	Answer the four open decisions	none	human, one sitting
C	Give the searched-surfaces listing a mechanical floor, or withdraw the claim	B2	agent-ready once B answered
D	Build sweep resume semantics: per-sweep fetch ceiling and cache, ledger enforcement, checkout-local	none	agent-ready, scoped
E	Re-score the golden set against rubric v3	B3, B4	agent-ready once B answered
F	Grow the golden set past `min_n_per_class`	E, #3458	tracked in #3458
G	Disposition-applying sweep	B1, F	blocked
H	Convention engagement as one changelog entry	G	blocked
I	Narrow CI regression gate for the cleaned state	G	blocked

A and D are executable today and depend on nothing in the decision set. If the maintainer wants movement before sitting down with B, those two are where it exists.

Routing

needs-human, for a genuinely open decision rather than a capability blocker: the four questions in B are spec calls about what the sweep's disposition scope is and what the tier table contains. Neither the decision-defaulted exit nor a delegable one fits, because there is no reversible default available for a question that redefines what the deliverable is.

The item stays as the sub-topic's Brief. Nothing here supersedes the stage-1 run comment, which remains the record of what was measured.

Labels: replacing the raw priority: needs-triage with priority: medium (real value, no deadline, and the blocked stages cannot move until B is answered regardless of urgency); adding needs-human.

Triage claim released.

Generated by Claude Code

This was generated by AI while splitting waiting issues into lane-ready slices (owner decisions C12, Q93-Q95, 2026-10-24).

Split into three dependent sub-issues, in this order:

attribution: run the report-only provenance sweep on current main and post its telemetry #9753: run the report-only sweep on current main under copy rubric v5 and post its telemetry here. Nothing blocks it.
attribution: run the disposition-applying provenance sweep once a class is fix-eligible #9755: the disposition-applying sweep. Blocked by attribution: owner rulings on blocked provenance panels and verbatim target #5654 and attribution: run the report-only provenance sweep on current main and post its telemetry #9753. If score-golden.sh still finds no fix-eligible class, it posts that result and stops.
upstream-drift: record the provenance convention entry and add the cleaned-state CI gate #9759: the upstream-drift convention entry, the hash-store trigger evaluation, the dead-pointer wiring and the cleaned-state CI gate. Blocked by attribution: run the disposition-applying provenance sweep once a class is fix-eligible #9755.

The sweep ledger is checkout-local, so slices 1 and 2 post their numbers as comments on this issue, and slice 3 reads them from there. The engagement text this Brief cited from docs/specs/provenance-convention-engagement.md (deleted by #5719) is carried from this issue's body into #9759.

Not a slice: #5654 holds the owner rulings, and it is parked under the no-eval-spend decision of 2026-10-21. Verbatim stays at n 3 against a target of 10, so only near-verbatim can become fix-eligible, and it was at 0.92 against a 0.95 bar on 2026-10-04.

The triage sweep dispatched intake triage for this issue (attempt 1 of 2).

Proposed work class: work-class: structural
Basis: automated intake triage; the lanes App applied this class.

Size: Large
Complexity: High
Uncertainty: Knowable
Flags: none
Blocked by: none

Understood: this is the Brief for the provenance sweep sub-topic. It covers a repo-wide report-only sweep, a disposition-applying sweep once a class becomes fix-eligible, the upstream-drift convention entry, and a cleaned-state CI gate. It has already been split into #9753, #9755 and #9759, so this issue now works as the tracking parent. The slices carry the actual dependencies, and #9755 depends on #5654 and #9753. I found no open question the issue text leaves unanswered and no likely duplicates.

_{Automated triage, applied by the lanes App.}

Not admitted to the issue lane yet: sub-issue #9755 is open; sub-issue #9759 is open.

Report-only sweep, 2026-10-28, `8dcdac6`

Slice 1 of 3 for this issue, from #9753. This ran /attribution:audit in its read-only audit action, with no fix and no sweep, over tracked markdown on main at 8dcdac6e8d93171f8642b44d9aaa5635491e6b46, in a detached worktree. The plugin was attribution 0.13.0 under Claude Code 2.1.312. The run applied no disposition and edited no tracked file: git status --porcelain and git diff --stat origin/main...HEAD are both empty, and no PR was opened.

Rubric versions: copy rubric v5 and restated-fact rubric v1. Every finding below was graded under one of those two.

The effective config came from docs/conventions/attribution.yaml, with everything else at the bundled defaults:

excluded_paths: **/evals/fixtures/**, trigger_less_stamp_check: true, stamp_expiry_days: 180
judge_samples: 3, nomination_passes: 2, judge_lens_diversity: true, review_agents: 1, deep_research_on_exhaustion: false
corpus_fetch_ceiling: 200, separation thresholds min_containment 0.3 and min_span_words 15, passed explicitly on every fingerprint call

Telemetry

Measure	Value
Files considered	3832 tracked markdown
Files declined	235: 214 `/evals/fixtures/` (config carve-out), 21 `/vendor/` (built-in carve-out)
Files included (enumerated)	3597
Breadcrumb inventory	1741 directories; 10446 URLs, 11 fences, 3024 stamp lines, 6037 quote lines
Stamp candidates / parsed / declined	3055 / 3031 / 24 (8 month-name dates, 8 bare years, 8 year-month only)
Expired stamps (past 180 days)	2
Trigger-less-stamp findings	49 (87 on 2026-09-06)
Files model-read for nomination	118 of 3597 (see Honest limits)
Nominations	1702 raw across 2 passes; 862 after union and dedupe (565 nominated in both passes), in 111 of 118 files
Nomination classes (raw)	restated-fact 1002, paraphrase 272, summary 205, verbatim 126, near-verbatim 97
Source fetches attempted / succeeded / failed	200 / 163 / 37, against the 200 ceiling
Fetch failures by reason	unexpected-content-type 16, redirected-off-path 9, http-403 4, not-in-index 3, too-large 3, http-429 1, slug-collision 1
Identity check failed after a 200 read	1 (a university paper URL served an error page; caught at judging)
429 count	1 (one generic-profile host; no retry)
Fingerprint comparisons	609 case-source pairs. 104 fired the separation rule: 102 on the 15-word span floor alone, and 2 (the same upstream file at two refs) also on containment 0.339. Merged into 246 distinct above-rule spans
Candidates judged	45 (29 copy, 16 restated-fact), 3 blind judges each (135 grades)
Refutation passes	3 (one per unanimous restated-fact STANDS)
Copy review passes (`review_agents: 1`)	0 run: no copy candidate stood
Categorical carve-outs	Corpus: the 2 path carve-outs above. Judgment: restated-fact carve-out 1 (conforming pointer or record) on 3 candidates unanimously, and on 3 more by 1 of 3 judges. Copy carve-outs: 0 applied; judges weighed carve-out 2 and graded the criteria anyway, as the rubric directs
Accumulated per-instance suppressions	0. There are no `attribution` entries in `docs/conventions/attribution.yaml`, `.claude/attribution.json`, `docs/conventions/harness-config.yaml` `suppressions`, or `.claude/audit-pass.md`
Embedded imperatives in read material	15 reported, by judges and refuters; nominators reported none. All were in local files or fetched pages. Each was reported as data and changed no verdict

Findings

Deterministic: expired stamps (2)

Location	Stamp date	Days over
`docs/upstream/claude-com-fold-subagents.md:67`	2025-06-12	323
`plugins/work-items/skills/track/actions/add.md:116`	2026-04-08	23

Class: stamp. Rule attribution/audit/rule-stamp-expired, deterministic, relay-eligible. Rubric: none. These came from the deterministic check-stamps.sh and were not panel-graded.

Deterministic: trigger-less stamps (49)

Class: stamp. Rule attribution/audit/rule-trigger-less-stamp, deterministic, relay-eligible. Rubric: none. These came from the deterministic check-stamps.sh and were not panel-graded. Each location carries a dated stamp on a surface that states no recheck trigger. The stamp date is in parentheses.

.claude/mutation-testing.md:46 (2026-10-25); docs/adr/0039-keep-babysit-identity-keys-in-userconfig-and-move-repository-keys-to-the-cascade.md:79 (2026-10-07); docs/adr/0055-load-nothing-head-controlled-into-a-pipeline-skill-activity.md:65 (2026-10-04); docs/adr/0065-move-the-harness-ops-session-event-log-into-a-mod.md:17 (2026-10-04); docs/adr/0071-start-pr-unblock-from-a-push-to-main-dispatcher-checked-at-the-gate-and-the-broker.md:69 (2026-10-11); docs/adr/0076-keep-guest-use-of-claude-the-owners-own-use.md:16 (2026-10-16); docs/conventions/hook-precision/guard-precision.md:85 (2026-10-15); docs/conventions/invocation-context/README.md:32 (2026-09-28); docs/conventions/permission-rule-hygiene/CHANGELOG.md:83 (2026-08-31); docs/conventions/plugin-reconfiguration/README.md:33, :51 (2026-10-28), :78 (2026-09-27); plugins/ai-briefing/README.md:29 (2026-07-18); plugins/bugs/CHANGELOG.md:69 (2026-10-07); plugins/code-tidying/skills/audit-comment-residue/SKILL.md:43 (2026-09-03); plugins/code-tidying/skills/audit-comment-residue/reference/shapes.md:21 (2026-09-03); plugins/context-budget/skills/audit/reference/report.md:37 (2026-10-11); plugins/context7/skills/lookup/context/cli.md:5 (2026-10-11); plugins/education/CHANGELOG.md:152 (2026-10-07); plugins/evals/evals/reference-read-denied/prompt.md:2 (2026-10-02); plugins/evals/skills/plugin-eval/reference/sandboxes.md:40 (2026-10-25), :54 (2026-10-23); plugins/harness-ops/skills/control-claude-code/features/desktop/mod-harness/README.md:35 (2026-10-26); plugins/harness-ops/skills/control-claude-code/features/skills/invoke/README.md:38 (2026-10-26); plugins/improvement/CHANGELOG.md:57 (2026-10-07); plugins/instruction-placement/context/findings-artifact.md:127 (2026-10-27); plugins/mutation-testing/skills/principles/reference/theory.md:100 (2026-10-06); plugins/pixel-art/palettes/README.md:106, :111, :116, :121, :125, :126 (2026-09-28); plugins/planning/reference/variation-points.md:145 (2026-10-11); plugins/planning/skills/design/SKILL.md:408 (2026-10-11); plugins/planning/skills/interview/context/assumption-sweep.md:17 (2026-09-27); plugins/playbooks/skills/boris/reference/effort.md:3 (2026-10-11); plugins/playgrounds/README.md:39 (2026-09-01); plugins/songwriting/context/pat-pattison/research/lyric-melodic-roadmaps.md:29, :64, :180 (2026-08-11); plugins/songwriting/skills/suno/reference/suno-drift-audit-ledger.md:17 (2026-08-12), :22 (2026-08-11), :23 (2026-07-18); plugins/testing/skills/audit/evals/judge-calibration/calibration.md:399, :409 (2026-10-26); plugins/user-interface/CHANGELOG.md:207 (2026-10-07); plugins/user-interface/reference/animation/gestures.md:11 (2026-10-11); plugins/x/CHANGELOG.md:58 (2026-10-06).

Several of these hits are probably not verification stamps: changelog entries, an eval prompt, and palette rows. They are adjudication candidates (below), not suppressions.

Judged: restated-fact, relay-eligible (2)

Rule attribution/audit/rule-restated-upstream-fact, class restated-fact, restated-fact rubric v1. Each had a unanimous 3/3 STANDS panel and survived refutation. Both are report-only and never fix-eligible.

Location	Tier	Source (rung)	Evidence (quoted)
`docs/upstream/aihero-shipping-course.md:268-277`	`source-fetched-similar`	https://code.claude.com/docs/en/hooks (rung 1, identity checked: "Hooks reference")	R1/R2: "a PostToolUse payload carries `tool_name`, `tool_input`, and `tool_response`, and nothing in it names the invoking skill". R4: no hooks-reference or OTEL pointer, as-of date, or trigger near the passage; the file's only whole stamp covers the course and mattpocock/skills. The refuter found the fact has already drifted: the fetched page now lists `session_id`, `transcript_path`, `cwd`, `permission_mode`, `tool_use_id` and an optional `duration_ms`.
`plugins/machine-health/skills/audit/reference/windows/check-catalog.md:414-419`	`not-found` (searched: in-file and directory breadcrumbs, nominator candidate URLs; no Microsoft Task Scheduler page fetched)	none located	R1/R2: "`LastTaskResult` is 267011 (`SCHED_S_TASK_HAS_NOT_RUN`) ... 267014 (`SCHED_S_TASK_TERMINATED`)". R4: section 10 has no pointer, as-of date or recheck trigger; the nearest record (section 11) covers Defender exclusions.

Judged: restated-fact, human report only (2)

Location	Outcome	Why it stays off the relay
`docs/cloud-sessions.md:232-240`	Unanimous STANDS, then REFUTED on attack (e), a close call. `leave-with-reason`.	The refuter read the "Measured 2026-10-10 on Claude Code 2.1.296" paragraph at lines 242-258 as a whole record: a probe as the pointer, a date and version, and "Re-measure ... on a later `main`" as the trigger. Open questions: whether "a later `main`" meets the observability bar for a Claude Code behavior, and whether the record covers the `window_tokens x bytes_per_token x fraction` formula and the 4-bytes half of the range.
`plugins/discovery/skills/trace-intent/SKILL.md:55-60`	3/3 UNKNOWN on R4	The passage hands its fact to "the dated record" in `reference/parent-contract.md`, which was outside the packet. A reader of that file settles it.

Judged: copy candidates, all cleared (29)

Copy rubric v5. Each candidate was 3/3 CLEARED, so none is a finding and none is fix-eligible. These were the 29 above-rule spans of 28 or more matched words.

Location	Source	Cleared on
`docs/upstream/claude-code.md:175`	code.claude.com env-vars	C1-C4 FAIL: no matching span on the page
`docs/upstream/good-css.md:95, 100, 109-112, 116-118, 129-133, 140` (6)	vojtaholik/good-css `PRACTICES.md` at the pin	C3 FAIL: "Credits are upstream's `Borrowed from:` lines at the pin, without their links" heads the table, and the file carries a whole record
`docs/plugin-philosophy.md:1243-1244`; `plugins/harness-config/skills/audit-instructions/reference/criteria.md:197-201, 1497-1499, 2157-2159, 2397-2398, 2473-2477` (6)	platform.claude.com pages	C2/C3 FAIL: each matched span is a link label equal to the source's section title, mostly the pointer of a whole stamped record
`plugins/playbooks/reference/model-adaptation/{sonnet-5-5,opus-5-5,haiku-5-5,fable-5-1}.md` (14) and `plugins/harness-config/skills/audit-prompting-postures/reference/postures.md:14-20, 114` (2)	the model-specific prompting pages	Same: link labels inside whole records (pointer, as-of date, trigger)

Judged: restated-fact, cleared (12)

Each was cleared by 3/3 judges, by carve-out 1 or by an R-criterion FAIL:

Location	Cleared on
`plugins/disk-hygiene/skills/clean/reference/safety-model.md:196-204`, `plugins/rate-limit-guard/README.md:297-307`, `plugins/performance/skills/go-faster/reference/catalog/harness.md:8-19`	Carve-out 1, unanimous
`plugins/harness-memory/skills/audit/reference/official-guidance.md:41-44`, `plugins/docs-hygiene/skills/audit-progressive-disclosure/context/tier-model.md:40-50`, `plugins/writing/skills/be-concise/reference/sources.md:103-108`, `plugins/session-flow/skills/orchestrate/SKILL.md:207-213`	R4 FAIL: a whole record covers the fact
`plugins/harness-config/skills/audit-instructions/reference/criteria.md:1653-1663`, `plugins/context-guard/reference/reader-contract.md:466-473`, `docs/migration-playbook.md:1474-1481`	Carve-out 1 per one judge, R4 FAIL per the other two
`plugins/naming/skills/name-it-better/context/sources.md:42-48`	R2 FAIL (a published paper version). Its fetched source failed identity
`plugins/overengineering/context/scrutiny-method.md:98-104`	R2/R3 FAIL (a historical incident)

Above the separation rule, not judged (217 spans in 40 files)

These spans of 15-27 words fired the deterministic rule but received no panel. Going by the 29 judged spans, they are mostly link labels and section titles, but none was judged here. Heaviest files:

plugins/harness-config/skills/audit-instructions/reference/criteria.md (42)
plugins/harness-config/skills/audit-prompting-postures/reference/postures.md (22)
plugins/playbooks/reference/model-adaptation/opus-5-5.md (19), sonnet-5-5.md (17), haiku-5-5.md (11), fable-5-1.md (8)
docs/plugin-philosophy.md (9), docs/conventions/mod-authoring/README.md (9), plugins/playbooks/reference/prompt-caching.md (8)
docs/upstream/claudedevs-cost-performance.md (7), docs/upstream/good-css.md (5)
29 more files with 1-5 spans each

Adjudication candidates for the golden set (harvested, not converted)

Link-label spans over the 15-word floor. A section title used as link text fires the span rule (102 of 104 firing pairs fired on the span floor alone) and then clears on C2/C3 every time it is judged. Example case: plugins/playbooks/reference/model-adaptation/fable-5-1.md:119. This is a precision case for the separation rule, or a quote-stripper case for link text.
Copied credit lines under a scoped attribution line. docs/upstream/good-css.md:95-140 passes C1, C2 and C4 and clears only on C3, through a table-scoped attribution sentence. This is a C3 scope-matching boundary case.
Probe-and-date record with a repository-side trigger. docs/cloud-sessions.md:232-258 drew a unanimous STANDS that refutation reversed. Does "a later main" count as an observable trigger for a vendor behavior?
Carve-out 1 versus R4 FAIL. criteria.md:1653-1663, reader-contract.md:466-473 and migration-playbook.md:1474-1481 reached the same outcome by two different routes across lenses.
A record in a sibling file. plugins/discovery/skills/trace-intent/SKILL.md:55-60 drew 3/3 UNKNOWN because R4's record lives in another file.
A 200 response whose body is an error page. The identity failure behind naming/.../sources.md:42-48 showed up only at judging, which argues for checking identity at fetch time.
Trigger-less-stamp hits on non-stamp text. Changelog lines (plugins/bugs/CHANGELOG.md:69 and five more), an eval prompt (plugins/evals/evals/reference-read-denied/prompt.md:2) and palette rows (plugins/pixel-art/palettes/README.md:106-126).

Honest limits

Files model-read versus only enumerated: 118 of 3597 were model-read for nomination. The other 3479 were only enumerated and covered by the deterministic steps: stamp check and breadcrumb inventory over all 3597. The 118 were chosen by a fixed rule:
- all 38 files under docs/upstream/
- the 60 files with the most non-org external URLs, excluding CHANGELOG.md and .changes/
- 20 drawn at random (seed 9753) from the remainder, also excluding changelogs
Nomination passes run: 2, unioned and never intersected: one led with wording, one with restated facts. Partial reads:
- pass 1 read criteria.md (2676 lines) only in part; pass 2 read it in full
- pass 2 read two table files in plugins/ only in part
- one chunk got a third, duplicate pass by accident, and its nominations were unioned in
- the brief capped nominations at 12 per file, and nominators noted further passages in more_seen that were not recorded
judge_samples panels: 3 of 3 lens-diverse, blind judges for every one of the 45 judged candidates; the floor was met. Disclosures:
- each judge graded a packet of 6-8 candidates under neutral labels, not a single candidate per dispatch
- nominators ran on Sonnet; judges and refuters ran on Opus 5.5
What was not judged:
- 217 above-rule spans
- 186 of the 202 strong restated-fact candidates nominated by both passes (16 were sampled, one per file, seed 9753)
- all lower-strength nominations
Fetch ceiling: it bounded source resolution. 411 distinct copy-class URLs were nominated: 200 were fetched, 31 x.com URLs were not attempted, and 180 more sat beyond the ceiling. No budgeted search ran for passages without a breadcrumb. deep_research_on_exhaustion was off.
The sweep ledger was not used. This was audit, not sweep. The relay findings file (53 rows: 51 stamp rows and 2 restated-fact rows) sits in a gitignored .work/reviews/main/ in one checkout, so this comment is the durable record.

Cordoned to the After zero milestone by the issue-triage session; not closed. Both open sub-issues, #9755 and #9759, are already in After zero, and admission holds this parent until they close (admission comment names both). Leaving it at status: ready counted it as lane supply it cannot be. The work stays open and resumes with its children.

Posted by the issue-triage session. This was generated by AI.

provenance sweep sub-topic: repo-wide sweep, upstream-drift convention engagement, and the cleaned-state gate #3465

Description

Why this is an issue rather than docs/topics/provenance-sweep/PLAN.md

Precondition, and it is a hard one

Measured baseline entering the sweep

Execution contract (inherited from the parent Brief)

The two additions the parent plan requires this Brief to carry

The convention engagement (fires at sweep completion only)

Checklist

Related

Activity

kyle-sexton commented on Aug 28, 2026

Stage 1 report-only sweep: first run, recorded

Checklist movement

Telemetry

The adoption-bar evidence this issue asks for

Resume semantics are not implemented

Findings

T15 reproduced live, twice

The no-breadcrumb probe validated, and found a gap

Detector defects found by running it on itself

Adjudication candidates harvested

Honest limits on this run

Recommended before stage 2

kyle-sexton commented on Sep 6, 2026

kyle-sexton commented on Sep 6, 2026

Triage: verified, oversized, routed human-gated with a decomposition proposal

Re-measured on current main, --as-of 2026-09-06

Why this is oversized

Why it is not being decomposed in this pass

Proposed decomposition, for approval in one pass

Routing

38 remaining items

kyle-sexton commented on Oct 24, 2026

github-actions commented on Oct 25, 2026

github-actions commented on Oct 25, 2026

melodic-automation-lanes commented on Oct 25, 2026

kyle-sexton commented on Oct 28, 2026

Report-only sweep, 2026-10-28, 8dcdac6

Telemetry

Findings

Deterministic: expired stamps (2)

Deterministic: trigger-less stamps (49)

Judged: restated-fact, relay-eligible (2)

Judged: restated-fact, human report only (2)

Judged: copy candidates, all cleared (29)

Judged: restated-fact, cleared (12)

Above the separation rule, not judged (217 spans in 40 files)

Adjudication candidates for the golden set (harvested, not converted)

Honest limits

kyle-sexton commented on Nov 2, 2026

Metadata

Metadata

Assignees

Labels

Type

Fields

Projects

Milestone

Relationships

Development

Issue actions

Why this is an issue rather than `docs/topics/provenance-sweep/PLAN.md`

Re-measured on current `main`, `--as-of 2026-09-06`

Report-only sweep, 2026-10-28, `8dcdac6`