PostToolUse
01Hook Claude Code
Runs after a tool call finishes for Edit and Write tool calls, running an inline shell check. From MiaoY0uShan/FP.
3 25d ago A
tokens not measured
copy · 100% MIT
SessionStart
02Hook Claude Code
Runs when a session starts, running an inline shell check. From MiaoY0uShan/FP.
3 25d ago A
tokens not measured
copy · 100% MIT
FP
03Settings file Claude Code
Agent settings declaring 2 hook events (PostToolUse, SessionStart).
3 25d ago A
tokens not measured
copy · 100% MIT
debug-issue
04Skill Claude CodeCodex
Systematically debug issues using graph-powered code navigation.
3 25d ago A 13 tokens
copy · 100% MIT
Skill Claude CodeCodex
Navigate and understand codebase structure using the knowledge graph.
3 25d ago A 15 tokens
copy · 100% MIT
Skill Claude CodeCodex
Plan and execute safe refactoring using dependency analysis.
3 25d ago A 15 tokens
copy · 100% MIT
Skill Claude CodeCodex
Perform a structured code review using change detection and impact.
3 25d ago A 14 tokens
copy · 100% MIT
MCP server Claude CodeCodexCursor +2
Local-first knowledge graph for token-efficient code review through MCP and CLI. Runs locally from the code-review-graph Python package.
3 25d ago A
tokens not measured
copy · 92% MIT
fp-host
09Skill Claude CodeCodex
Host-native FP variant for agent CLIs with built-in planning, subagents, progress UI, and session resume (Claude Code, Codex CLI, and similar). Keeps the four core rules, routing judgment, and domain profiles; delegates orchestration machinery to the host. Experimental: not yet blind-eval validated — use fp/SKILL.md…
3 25d ago A 77 tokens
original MIT
fp-minimal
10Skill Claude CodeCodex
Minimal FP variant for non-reasoning models. 4 rules, no ceremony.
3 25d ago A 21 tokens
original MIT
fp
11Skill Claude CodeCodex
Use automatically when the user's goal is engineering work (build, change, diagnose, review, test, operate, or plan software, repositories, infrastructure, or agent tooling), or when explicitly invoked with "FP:" or "$fp". Do not use for casual conversation or other non-engineering goals.
3 25d ago A 61 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP after a non-trivial evidenced run to stage a reusable checklist, schema, or automation candidate.
3 25d ago A 31 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP after an executable plan is accepted and the current host exposes a verified subagent runtime; dispatch a fresh implementer, task reviewer, bounded fixer/re-review loop, final integration reviewer, and parent verification.
3 25d ago A 53 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP after requirements work to apply the seven-rung reuse ladder, delete scope, and define the smallest verifiable MVP.
3 25d ago A 34 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP for two or more explicitly independent problem domains whose dependencies, writer paths, generated outputs, state, and verification can be isolated before bounded fan-out and parent fan-in.
3 25d ago A 47 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP after execution, failure, or handoff to record canonical claims, checks, scope, authority, experiments, and optional continuation evidence.
3 25d ago A 38 tokens
original MIT
Skill Claude CodeCodex
Evaluate evidence-backed checklist, schema, skill-patch, or automation candidates before shadowing or promotion. Use after adaptive improvement, for background-learning proposals, or whenever a rule inferred from a small number of runs could overfit, underfit, leak its holdout cases, or cause negative transfer.
3 25d ago A 64 tokens
original MIT
fp-metrics
18Skill Claude CodeCodex
Use only when routed by FP after a canonical Evidence Ledger and when measurement or a comparable baseline matters.
3 25d ago A 24 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP after MVP scope is defined to choose the first safe reuse rung and compile the final Execution Brief.
3 25d ago A 30 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP for an agent host using a non-native model provider, API-compatible endpoint, gateway, or local proxy, or when loops, retries, token spend, cache accounting, streaming, tool calls, or encoding are suspect.
3 25d ago A 55 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP for vague, risky, large, product-facing, or assumption-heavy work; separates facts, decisions, failure paths, success criteria, and non-goals.
3 25d ago A 43 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP after repeated evidence has accumulated; store reusable patterns and stop conditions, never raw context.
3 25d ago A 27 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP after delete-scope and only when the scoped MVP crosses modules or has real coupling risk.
3 25d ago A 30 tokens
original MIT
Skill Claude CodeCodex
Use only when routed by FP because work is too large, an attempt failed, or the selected path needs evidence-backed micro-loops.
3 25d ago A 33 tokens
original MIT