Verification trail
#4
✅ Verified
.claude/commands/docs-review/scripts/test_apply_update.py:554
assert "https://github.com/pulumi/docs/blob/master/CONTRIBUTING.md#ai-assisted-contributions" in _footer(a2)
evidence: CONTRIBUTING.md in the repo has a heading at L17, "## AI-assisted contributions". GitHub turns that into the anchor #ai-assisted-contributions, so the URL in the test assertion points at a real section.; source: repo:CONTRIBUTING.md#L17
source: repo:CONTRIBUTING.md#L17
pass1 · url · high confidence
#1
➖ Not a claim
.claude/commands/docs-review/references/output-format.md:340
- **Editorial stances (no verdict).** `compose-review.py` renders a second H4 inside ⚠️ Low-confidence, **before** `#### Style suggestions`: `#### Editorial…
evidence: This line describes how the PR author's own review tooling works: the order in which `compose-review.py` renders headings, and the `stances` list in `.candidate-claims.json`. The regex matched the word "before" because it looked…; source: repo:.claude/commands/docs-review/references/output-format.md L340
source: repo:.claude/commands/docs-review/references/output-format.md L340
pass1 · temporal · high confidence
#2
➖ Not a claim
.claude/commands/docs-review/scripts/compose-review.py:2176
return (f"https://github.com/{repo}/blob/master/CONTRIBUTING.md#ai-assisted-contributions"
framing: Templated URL in code; the {repo} value is chosen at runtime, not stated as a fact.; evidence: This is an f-string URL template inside the PR author's own review-tooling script. The `{repo}` placeholder is filled in at runtime, so the line is the author's own design, not a statement about a third party. The target it points to…; source: WebSearch ran query "pulumi docs CONTRIBUTING.md \"AI-assisted contributions\""; https://github.com/pulumi/docs/pull/21901
source: WebSearch ran query "pulumi docs CONTRIBUTING.md \"AI-assisted contributions\""; https://github.com/pulumi/docs/pull/21901
framing: Templated URL in code; the {repo} value is chosen at runtime, not stated as a fact.
pass3 · url · medium confidence
#3
➖ Not a claim
.claude/commands/docs-review/scripts/compose-review.py:2183
# above needs one of these" (reader feedback, 2026-09-25).
framing: This is an internal provenance note in a code comment. It doesn't cite any external source.; evidence: The line is a code comment in the PR author's own review-tooling script. It records the origin of a design decision ("reader feedback, 2026-09-25"). The regex flagged the date as numerical, but the line makes no factual claim about a…; source: .claude/commands/docs-review/scripts/compose-review.py:L2183
source: .claude/commands/docs-review/scripts/compose-review.py:L2183
framing: This is an internal provenance note in a code comment. It doesn't cite any external source.
pass3 · numerical · high confidence
#5
➖ Not a claim
.claude/commands/docs-review/scripts/testdata/v3-fixture-author.md.txt:48
*Optional polish from pattern-based linting and the review's own read; never blocking. ✏️ marks one you can apply from the [Files…
framing: Test fixture content; the URL is a sample PR link and makes no factual claim.; evidence: This line comes from a test fixture (`scripts/testdata/v3-fixture-author.md.txt`) for the docs-review tool. The link to `pulumi/docs/pull/999/files` is a made-up PR number used in sample review output, not a documentation claim that…; source: repo:.claude/commands/docs-review/scripts/testdata/v3-fixture-author.md.txt
source: repo:.claude/commands/docs-review/scripts/testdata/v3-fixture-author.md.txt
framing: Test fixture content; the URL is a sample PR link and makes no factual claim.
pass1 · url · high confidence
No claims match the current filter.
Findings
No findings.
Editorial stances
Superlative, ranking, or comparative language the diff adds. No verdict — a page's own framing isn't fact-checkable — listed so a reviewer can confirm each is a stance the docs should take.
- .claude/commands/docs-review/references/output-format.md L340 - **Editorial stances (no verdict).** `compose-review.py` renders a second H4 inside ⚠️ Low-confidence, **before** `#### Style suggestions`: `#### Editorial… comparison · found by regex
- .claude/commands/docs-review/references/output-format.md L340 - **Editorial stances (no verdict).** `compose-review.py` renders a second H4 inside ⚠️ Low-confidence, **before** `#### Style suggestions`: `#### Editorial… positioning · found by regex
Investigation log
- cross-sibling-reads
- not run (not in a templated section)
- external-claim-verification
- 1 of 5 claims verified (0 unverifiable, 0 contradicted) · 4 specialists (numerical, cross-reference, capability, framing); 0 cross-specialist corroborations · routed: 0 inline, 3 Pass 1, 0 Pass 2, 2 Pass 3 (verified 0, contradicted 0, unverifiable 2).
- cited-claim-spot-checks
- not run (no cited claims)
- frontmatter-sweep
- not run (no frontmatter in diff)
- temporal-trigger-sweep
- ran (recency words present in diff; spot-check in-review)
- code-execution
- not run (no `static/programs/` change)
- code-examples-checks
- not run (no fenced code blocks in content files)
- editorial-balance-pass
- not run (not under content/blog/)
History
- 2026-09-26T00:44:23Z
3a6b06f
initial review (pending publication)