agentleFS
zedarvateszedarvates/botte-secrete/skills/asset_quality/SKILL.md

asset-quality

Skill2 starsChanged 3 months agoMITView on GitHub

Evaluate generated images, textures, meshes, animations, and Godot packages with deterministic gates followed by an explainable family-local k-NN memory.

The file

Files
SKILL.md

Asset Quality Memory

Asset Quality Memory is a zero-cloud, standard-library baseline for Asset Factory QA. It is deliberately shadow-only: it explains FAIL, UNCERTAIN, PASS, or PASS_ROBUST, but never imports, publishes, or activates an asset.

Decision order

  1. Decode, manifest, SHA-256/size, and licence checks.
  2. Family-specific deterministic checks.
  3. k-NN over externally verified assets from the same family only.
  4. Abstention when fewer than three comparable examples exist.
  5. A future micro-NN is considered only after beating this baseline on a temporal holdout without reducing quality.

Supported families are image, texture, mesh, animation, and godot. Their feature schemas are intentionally separate; a mesh can never become a neighbor of an image.

Commands

botte asset-qa status . --json
botte asset-qa evaluate asset-report.json --project . --json
botte asset-qa record asset-report.json --project . \
  --verdict pass --verified-by tests:asset-harness --evidence ci:asset-142

The report contains family, sha256, size_bytes, checks, and normalized features. See the JSON schema and examples/asset-quality/mesh-report.json for a complete input. Operational memory is written to .botte/asset-quality.jsonl; local paths and raw asset bytes are never stored.

This implementation runs on CPU and mono-GPU machines. An RTX 3060 or other GPU may generate embeddings upstream later, but it is not required or trusted by the integrity gates.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.