v1.1.2

What's Changed

• Give the Pages site real metadata, and put claim_auditor back on it by @AlveeeRahman in #2
• Put the hallucination-hunter agent where it can actually be found by @AlveeeRahman in #3
• Fit the description inside claude.ai's 200-character limit by @AlveeeRahman in #4
• Release v1.1.2 by @AlveeeRahman in #5

New Contributors

• @AlveeeRahman made their first contribution in #2

Full Changelog: v1.1.1...v1.1.2

v1.1.1

Full Changelog: v1.1.0...v1.1.1

Skill Vision v1.1.0 — see the codebase, not just the verdict

A sixth tool, alongside the five validators: skill_mapper.py draws a skill's files and reference graph as a Mermaid flowchart, instead of a score.

Why: progressive disclosure means Claude starts at SKILL.md and follows links outward. A file nothing links to is never read, however well it's written — that's a shape problem, and a picture answers it faster than a findings list.

What's new

• scripts/skill_mapper.py — Mermaid flowchart (TD/LR/BT/RL), grouped by directory. --detail adds each script's function/class counts and CLI flags read straight off its AST. --json for the graph as data. No call-graph inference: published recall for static Python call graphs tops out around 70%, so this tool draws only what's checkable against ls and grep — files and resolved links, nothing guessed.
• Same resolver as the spec check, on purpose. spec_validator.py's link-checking logic is now build_graph(), a shared function both tools call. The map and the CONFORMANT/NOT CONFORMANT verdict use the same graph, so they can't drift apart.
• 131 tests (up from 117), including a parity suite asserting the mapper and the spec validator report the same broken and orphaned files on the same skill.
• Writes no file by default, only Mermaid to stdout — pipe it yourself: python3 scripts/skill_mapper.py path/to/skill --fence > diagram.md.

Install / upgrade

git clone https://github.com/AlveeeRahman/skill-vision.git ~/.claude/skills/skill-vision
# or, for an existing clone:
git -C ~/.claude/skills/skill-vision pull

Docs: https://alveeerahman.github.io/skill-vision/

Skill Vision v1.0.0 — the Claude skill evaluator and scorer

The QA skill for Claude skills: install it, then ask Claude to validate, test, and score your other Agent Skills before they ship.

The launch story: we ran the five validators over a private 20-skill corpus (145 scripts, ~45k LOC). The audit's most valuable output was 13 defects in this harness itself — every critical one is fixed in this release, each with a regression test. A validator that audits itself honestly is the whole point.

What's in v1.0.0

• spec_validator — Agent Skills spec conformance (loads? uploads? triggers?) + per-skill token-cost report (description tokens load every session, body on trigger)
• skill_validator — structure and documentation
• script_tester — executes every bundled script, recursively, with crash-honest verdicts
• quality_scorer — five-dimension score, A+ to F, improvement roadmap
• security_scorer — credentials/injection/traversal posture, false-positive-hardened (anchored patterns, docstring-aware)
• 117 adversarial self-tests, CI on Python 3.9–3.13, zero dependencies (pure stdlib — runs in the no-network Claude API container)

Install

git clone https://github.com/AlveeeRahman/skill-vision.git ~/.claude/skills/skill-vision

Docs: https://alveeerahman.github.io/skill-vision/ · Post your best skill's score in a checkup issue.