ClaimCheck What it does
ClaimCheck is an installable skill for AI agents. Flag literacy gaps in claims and arguments.
Give it an article, post, assertion, or short argument. An agent reads the skill, runs seven literacy dimensions, and writes a report. Each finding quotes the claim or names what is missing, gets one status, says why, and names the primary source to check next.
Get started: install the skill in your agent, then use it.
What it reads, writes, and changes
| Reads | An article, post, assertion, thread, or short argument |
| Writes | <name>.claimcheck.md next to a file, or claimcheck.md for a URL or pasted text |
| Changes | Nothing. The source is never edited |
What it does not do
- Rule a claim true or false. It classifies seven dimensions.
- Rewrite the claim, unless you ask separately.
- Review tone or phrasing. That is Smell Check.
- Stress-test a plan you are still deciding. That is TemperPass.
- Fill in a number it could not check. That becomes
unverified, with the source to open.
When to use which
| You have | Use |
|---|---|
| A claim, post, or argument | ClaimCheck |
| Prose that reads padded, loud, or off-register | Smell Check |
| A plan or decision on the table | TemperPass |
| A finished README, skill, or site | Cold-eye |
| Files after edits | Detangler |
Like Cold-eye and Detangler, ClaimCheck is non-destructive: a report next to the subject, the subject left alone.
The dimensions. The report shape.