coddy-agent / rules
Cursor rule187 starsChanged 9 days agoMITView on GitHub
Skill execution semantics and bundled skill delivery
Files
--- description: Skill execution semantics and bundled skill delivery globs: - "internal/skills/**/*.go" - "internal/agent/plugin_command.go" - "internal/tools/**/*skill*.go" - "external/ui/src/ui/skills/**/*" - "docs/features/skills.md" - "docs/reference/slash-commands.md" - "scripts/bundled-skills.json" - "scripts/vendor-bundled-skills.sh" - "internal/skills/bundled/**/*" alwaysApply: false --- # Skills and bundled delivery - A skill is instruction content for the current agent, never a hidden child run or `context:fork`. - Isolated execution uses an explicit subagent definition and `spawn_agent`. - Handle context size with concise skills, progressive disclosure, references, delegation, and compaction. - Bundled skills are embedded and seeded locally; startup never clones upstream. Keep `//go:embed all:bundled` so dot-directories remain included. - Tests that are not about the system marketplace use the offline `SystemSources` helper. - Replace an installed bundled skill only when the bundled version is newer. A missing version is legacy; preserve same-version, newer, or unreadable user copies. - Vendor through the established vendor and check targets, not manual generated-file edits.
In this repository31 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .cursor/rules/acp-protocol.mdcCursor rule
- .cursor/rules/agent-model-choice.mdcCursor rule
- .cursor/rules/api-layer.mdcCursor rule
- .cursor/rules/architecture.mdcCursor rule
- .cursor/rules/build-tags.mdcCursor rule
- .cursor/rules/code-style.mdcCursor rule
- .cursor/rules/config-roundtrip.mdcCursor rule
- .cursor/rules/console-tui.mdcCursor rule
- .cursor/rules/core-modules.mdcCursor rule
- .cursor/rules/docs-translation.mdcCursor rule
- .cursor/rules/gateway.mdcCursor rule
- .cursor/rules/implementation-order.mdcCursor rule
- .cursor/rules/llm-streaming.mdcCursor rule
- .cursor/rules/logger-components.mdcCursor rule
- .cursor/rules/merged-branch-cleanup.mdcCursor rule
- .cursor/rules/process-lifecycle.mdcCursor rule
- .cursor/rules/provider-auth.mdcCursor rule
- .cursor/rules/provider-devin.mdcCursor rule
- .cursor/rules/provider-proxy.mdcCursor rule
- .cursor/rules/russian-wording.mdcCursor rule
- .cursor/rules/session-storage.mdcCursor rule
- .cursor/rules/testing.mdcCursor rule
- .cursor/rules/text-encoding.mdcCursor rule
- .cursor/rules/tool-images.mdcCursor rule
- .cursor/rules/turn-admission.mdcCursor rule
- .cursor/rules/ui-assets.mdcCursor rule
- .cursor/rules/ui-spa.mdcCursor rule
- .cursor/rules/ui-verification.mdcCursor rule
- .cursor/rules/workflow.mdcCursor rule
- .cursor/rules/workspace-identity.mdcCursor rule
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
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.