Pagekit claim check
Run a skeptical claim-check pass on a PageKit first-page draft, at light/normal/hard severity. Step 07 of the PageKit method. Use after a first-page draft exists and you want to validate it before treating it as publishable. Delegates to the pagekit-claim-checker subagent.From its SKILL.md
npx -y skills add hnshah/pagekit --skill pagekit-claim-checkAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 10 stars10 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
2.8 KB, as published. Nobody here has run it
PageKit step 07 — Claim Check
You are running the optional durability pass on a first-page draft.
Read first
./references/prompt.md./references/framework.md— claim-check flag types, severity calibration, principle./references/template.md../pagekit/references/anti-slop.md— the slop pattern catalog
Inputs
runs/<run-name>/first-page-draft.mdruns/<run-name>/proof-map.md- Severity: light, normal, or hard. Ask the user if not specified; default to
normal.
Procedure
Option A — delegate to the subagent (preferred)
Invoke the pagekit-claim-checker subagent (agents/pagekit-claim-checker.md). Pass:
- the path to the draft
- the path to the proof map
- the severity
The subagent runs read-only, produces the audit, and writes:
runs/<run-name>/claim-check.md— the audit (flagged lines, types, rewrites, recurring patterns)runs/<run-name>/first-page-draft-corrected.md— the corrected draft (only if any line was rewritten)
Option B — run inline
Use scripts/claim-check.sh <draft> <proof-map> --severity <severity> to expand the canonical prompt for paste. Save outputs to the same paths.
After the pass
- Save the audit and corrected draft.
- Update
working-log.mdwith the severity used, count of flags by type, and any recurring patterns. - If the same slop pattern showed up more than once, add it to
../pagekit/references/anti-slop.md(or strengthen the existing entry) and to../pagekit-page-argument-shape/references/template.md's drafting-constraints list. The framework absorbs what the claim-check kept catching.
Quality gate
A strong claim-check pass:
- names specific lines, not general impressions
- separates claim types (overclaim / vague / unsupported / outrun proof map / editorial voice / unsourced quantitative / clinical drift / ai-slop tell)
- proposes replacements that are sharper and more accurate than the originals, not simply softer
- ends with a recurring-patterns summary so upstream constraints can absorb them
Next
Do not stop at claim-check. Continue:
- Write
runs/<name>/evaluation.md— the run's own honest read of what worked, what stayed thin, where outputs drifted generic. - Invoke
pagekit-evaluator-passto produce the adversarial second read. - Ensure
runs/<name>/working-log.mdhas real entries per step (replace all scaffold*[Fill in]*placeholders). - Invoke
pagekit-run-check— it must returntier: FULLY LOGGEDortier: PUBLISHABLEbefore the run is done.