generate-my-taste

Solid

Generate a personal taste skill from local evidence and confirmation forks.

AI & Automation 33 stars 0 forks Updated yesterday Apache-2.0

Install

View on GitHub

Quality Score: 86/100

Stars 20%
51
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
100
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

# generate-my-taste Generate a personal Claude Code taste skill from local evidence. The output is a `<name>-taste` skill shaped like `spine` / `taste`: frontmatter, posture, modes, two-sided charter, anchors, audit output shape, anchor mode output shape, auto-clarity exception. Evidence leads. Questions confirm. Generation writes only after a preview gate. ## Posture Derive taste from observed patterns, not self-report alone. Mine memories and local agent histories for repeated positive influences, Side A slop bans, and Side B overkill bans. Convert those signals into exactly 5 portable anchors unless the user explicitly selects update-in-place for an existing owned skill with a different anchor count. Keep the generator right-sized: exactly three generated files — `SKILL.md`, `references/anchors.md`, `references/charter.md`. ## Modes [LOAD-BEARING] ### Mode-selection This skill has one mode: generate. If invoked for audit of an artifact, route to an existing taste/spine skill instead. ### generate mode procedure 1. Resolve the target name. 2. Scan evidence sources. 3. Extract influence, slop, and overkill signals. 4. Ask compact confirmation forks. 5. Compose and show the synthesis preview. 6. Write generated references first and generated `SKILL.md` last. 7. Report written paths and verification notes. ## Evidence scan order Prefer indexed sources when available; fall back to files. 1. Indexed ICM tools or memory search tools exposed by the current harness. 2....

Details

Author
OutlineDriven
Repository
OutlineDriven/odin-claude-plugin
Created
8 months ago
Last Updated
yesterday
Language
Python
License
Apache-2.0

Bundled in these plugins

Similar Skills

Semantically similar based on skill content — not just same category