Help us improve
Share bugs, ideas, or general feedback.
From agent-harness-skills
Designs or refactors agent entrypoint files (AGENTS.md, CLAUDE.md, etc.) to establish source-of-truth navigation, mirror strategy, and drift prevention.
npx claudepluginhub yfge/codex-plugins --plugin agent-harness-skillsHow this skill is triggered — by the user, by Claude, or both
Slash command
/agent-harness-skills:agent-entrypoint-designThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Create a small, stable entrypoint that tells agents where truth lives, what rules apply, and which checks gate changes.
Creates p5.js generative art with seeded randomness, noise fields, and interactive parameter exploration. Use for algorithmic art, flow fields, or particle systems.
Share bugs, ideas, or general feedback.
Create a small, stable entrypoint that tells agents where truth lives, what rules apply, and which checks gate changes.
The entrypoint should be navigation plus hard rules, not a long project encyclopedia. For shared harness terms, see ../../references/harness-patterns.md; when entrypoint files are absent, use references/build-when-missing.md. For mirror casing and drift rules, see references/mirror-policy.md.
AGENTS.md, CLAUDE.md, GEMINI.md, .cursorrules, or GitHub instructions.AGENTS.md, CLAUDE.md, GEMINI.md, .cursor, and .github/instructions.references/build-when-missing.md.AGENTS.md in that subtree instead of overloading the root.# Agent Entrypoint Design
## Detected Mapping
- entrypoint:
- mirrors:
- validation:
## Source Of Truth
- Primary entrypoint:
- Mirrors:
- Subtree overrides:
## Root Entrypoint Sections
-
## Sync / Drift Prevention
-
## Minimum Commands
-
## Follow-up Files
-