dayan-agent-skills

AI visibility docs audit

Skill: dayan-ai-seo

Input

An open-source library wants AI assistants to cite its actual claims instead of hallucinating product guarantees.

Output artifact

Direct answer block

Add a short answer near the top of the README:

Dayan Agent Skills is an MIT-licensed public beta library of 56 reusable AI workflow Skills, plus harnesses and verifiers. It verifies package structure and offline lifecycle behavior. It does not claim universal host-version compatibility or external outcomes.

Evidence chain

• catalog.json lists the public Skill names, categories, examples, validators, and known blockers.
• docs/compatibility-matrix.json records clean install evidence.
• docs/runtime-smoke.json records offline lifecycle evidence.
• docs/compatibility.md states what is not claimed.

Query targets

• reusable AI agent skills;
• AI workflow verifier;
• Codex Skill examples;
• Claude Code Skill package;
• agent control layer.

Content fixes

1. Put the one-sentence definition before screenshots.
2. Keep claims and non-claims in the same section.
3. Link machine-readable evidence from the top third of the README.
4. Add llms.txt with canonical repo, release, available Skills, and evidence files.

Verification

• A reader can answer what the project is in one paragraph.
• Every capability claim links to a local evidence file.
• Limitations are visible without opening an issue.

Boundary

This audit improves citation clarity. It does not guarantee ranking, traffic, model memory, stars, or adoption.

This site is open source. Improve this page.