AI-assisted coding setup that helps you spend less time re-explaining context, and more time shipping with better output. Includes a custom spec kit, memory, agent and skill framework.
Runs before the agent uses a tool for Bash, Task, Task, Agent, mcp__claude_ai_.*, Write and Edit tool calls, executing dispatch-preflight-lint.mjs, spec-gate-enforce.mjs, git-preflight-advisory.mjs, task-dispatch-guard.cjs, fable-subagent-guard.mjs and mcp-route-guard.cjs via bash (7 commands). From…
Runs when you submit a prompt, before the agent sees it, executing user-prompt-submit.js and spec-gate-classify.mjs via bash (2 commands). From MichelKerkmeester/skilled-agent-harness_spec-driven-loops.
Runs when a session starts, executing session-prime.js, worktree-guard.sh, check-git-hooks.sh, git-primary-reconcile.sh, git-live-follow.sh, check-dist-staleness.sh and install-codex-hooks.mjs via bash with --start, --all and --check (7 commands). From MichelKerkmeester/skilled-agent-harness_spec-driven-loops.
Runs when the agent finishes a response, executing session-stop.js and completion-evidence-stop.cjs via bash (2 commands). From MichelKerkmeester/skilled-agent-harness_spec-driven-loops.
Runs after a tool call finishes for Write, Edit and Bash tool calls, executing claude-posttooluse.cjs and dispatch-audit-posttooluse.mjs via bash (2 commands). From MichelKerkmeester/skilled-agent-harness_spec-driven-loops.
★not rated 34 todayA
tokens not measured
copy · 100%MIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: