ai planning skills

azalio/map-framework

Skill Claude CodeCodex

ARCHITECT phase - decompose complex tasks into atomic subtasks with research, spec, and branch-scoped plan artifacts under .map.

azalio/map-framework

Skill Claude CodeCodex

Interactive 4-section code review using monitor, predictor, and evaluator agents plus the user and maintainer role reviewers on current changes. Use when reviewing a diff, PR, or staged work before merge. Do NOT use to plan or implement; use map-plan or map-efficient.

azalio/map-framework

Skill Claude CodeCodex

TDD MAP workflow: write tests from the spec FIRST, then implement, so tests validate intent not implementation. Use when correctness is critical (auth, payments, data integrity). Do NOT use without a spec; use map-efficient instead.

primeline-ai/universal-planning-framework

Skill Claude CodeCodex

The Universal Planning Framework rulebook - 10-step planning pipeline, 8 domain profiles, 21 anti-patterns with detection rules, mandatory FAILED conditions, and the C/B/A quality rubric. Use when creating, reviewing, refining or grading any plan, and whenever a plan needs discovery, kill criteria, observable gates or…

primeline-ai/universal-planning-framework

Skill Claude CodeCodex

The execution-vehicle rubric - decides silently how each plan phase runs (single agent, parallel sub-agents, agent team, background session, dynamic workflow or goal-loop) and which model tier it uses. Use while drafting or reviewing plan phases; never prompt the user for a vehicle.

ma-ziwei/fathom-mode-for-claude

Skill Claude CodeCodex

Enter Fathom Mode — a planning session where Claude pauses execution and helps the user build shared understanding through structured dialogue. Activate this whenever the user says "fathom", "fathom mode", "let's plan this", or asks to think through a complex task before acting. Stay in Fathom Mode until the user…

ma-ziwei/fathom-mode-for-claude

Skill Claude CodeCodex

Enter Fathom Mode — a planning session where Claude pauses execution and helps the user build shared understanding through structured dialogue. Activate this whenever the user says "fathom", "fathom mode", "let's plan this", or asks to think through a complex task before acting. Stay in Fathom Mode until the user…

paulbreuler/limps

Skill Claude CodeCodex

Guide for creating well-structured, meaningful Git commits following conventional commits and repository best practices. Use when creating commits, amending commit messages, or reviewing commit history.

paulbreuler/limps

Skill Claude CodeCodex

Common operations and planning principles for limps feature plans. Covers plan identification, artifact loading, distillation rules, and lifecycle guidance using limps MCP tools.

paulbreuler/limps

Skill Claude CodeCodex

Review code for correctness, security, and LLM/MCP safety in this repo. Use when a code review, security review, or MCP/LLM audit is requested.

kbichave/skills

Skill Claude CodeCodex

Standalone multi-expert code review, outside of /deep implement. Routes the diff to a panel of specialist reviewer subagents (core packs, logic, architecture, ML, stats, MLOps, data engineering, prompt engineering), verifies claims against current docs, gates everything through a final review-verifier, and writes a…

kbichave/skills

Skill Claude CodeCodex

Discovery, planning, and implementation pipeline. Modes — discovery: system audit → phase specs; plan: implementation blueprint; implement: execute sections; auto: autonomous end-to-end. Accepts @path, inline text, or no argument.

kbichave/skills

Skill Claude CodeCodex

Remove signs of AI-generated writing from text. TRIGGER when user says "humanize", "make this sound human", "remove AI writing", "de-slop", or asks to edit/review text to sound more natural. Also triggers on slide decks, roadmap docs, card headers and chart labels, not only prose. Based on Wikipedia's "Signs of AI…