Skills, prompts, AGENTS.md files and MCP servers that Codex reads. Skills are shared with Claude Code; the rest is Codex-native. 274,319 indexed, measured the same four ways as everything else here.
Skill Codex
Self-review gate and outcome decision checklist for BitGN PAC tasks. Use IMMEDIATELY before calling bitgn answer to verify correctness, security, outcome code, and refs completeness.
Skill Claude CodeCodex
High-seniority project manager agent. Plans work, decomposes requests into high-quality tasks with clear acceptance criteria, selects the next unblocked task for other agents, and validates completion vs requirements. Only Project-Manager defines/edits task entries; other agents must not create tasks.
Skill Claude CodeCodex
Proactive codebase analysis agent that scans for the highest-impact improvement opportunities across architecture, security, performance, testing, code quality, and more. Use when asked to "look for work", audit a codebase, or find what to fix next.
WanfengzzZ/harness-engineering
Skill Claude CodeCodex
A software-engineering approach for preparing a code repository so an AI agent can work through tasks independently. It combines clear rules, useful documentation, automated feedback such as tests and CI, and checks that verify the result.
Skill Claude CodeCodex
Push committed work, open a PR against dev, and land it with a merge commit so no commit is ever rewritten or duplicated. Use when invoked as /iso-push [--pr] [--cascade test|prod] [--no-merge] [--stay], or when asked to push a branch, open a PR against dev, or promote dev to test or prod. Requires a dev or develop…
Skill Codex
Use when a Codex repository relies on OpenSpec proposals and the work is moving from approved change to implementation in an isolated git worktree. Covers bootstrapping repositories with owf init, inspecting lifecycle state, starting implementation worktrees, applying migration rules, and cleaning up after merge. Do…
jidechao/legacy-context-builder
Skill Claude CodeCodex
A workflow for documenting and organizing older or hard-to-understand software so AI coding agents can maintain it safely.
Skill Claude CodeCodex
A skill that turns conversations or user actions in Tree Monstor into a new skill file in the project’s skills folder.
Skill Claude CodeCodex
Pick the app profile (A=Python / B=HTML+JS) via a plain-language decision tree and set the spec flags. Trigger: starting a new app in sdd.schema: reasons mode, or when the profile is undecided.
Skill Claude CodeCodex
A design and implementation guide for managing an order’s lifecycle, such as moving from pending to paid, shipped, and delivered. It also covers recording status history and preventing conflicting updates.
Skill Claude CodeCodex
Research the project and generate a holistic AGENTS.md file at the repo root. Use when the user says "init agents.md", "create AGENTS.md", or the project lacks agent instructions.
Skill Claude CodeCodex
Generate copy-paste bash scripts for Ralph Wiggum/AI agent loops (Codex, Claude Code, OpenCode, Goose). Use when asked for a "Ralph loop", "Ralph Wiggum loop", or an AI loop to plan/build code via PROMPT.md + AGENTS.md, SPECS, and IMPLEMENTATIONPLAN.md, including PLANNING vs BUILDING modes, backpressure, sandboxing…
Hhh2178/project-harness-governance
Skill Codex
Use when starting, restructuring, or maintaining a project repository that needs durable Codex guidance, clear document ownership, requirements and acceptance records, current-state tracking, verification contracts, and an auditable work log.
Skill Codex
Audit and govern local Codex architecture. Use when the user asks where to store or configure AGENTS.md rules, project knowledge, memories, skills, SKILL.md files, references, hooks, subagents, plugins, MCP servers, connectors, ECC assets, context management, or Codex architecture diagrams; when creating or updating…
Skill Claude CodeCodex
Cognitive memory skill for MuninnDB - store and retrieve memories with temporal priority, Hebbian learning, and automatic associations.
Skill Claude CodeCodex
A guide for working with Pluxel, including its plugin system, demos, HMR, MCP, runtime logs, tests, and TypeScript declaration files.
Skill Codex
Turn a repository into a compact, evidence-backed ontological index: a durable map of system purpose, containers, components, responsibilities, interfaces, data, and important code paths. Use the result to compress repository context and retrieve the source locations that explain how a part of the system works.
longranger2/project-doc-bootstrap
Skill Claude CodeCodex
Bootstrap a reusable repository documentation hierarchy for AI agents. Use when starting a new project, initializing repo docs, setting up CLAUDE.md and AGENTS.md, defining L1/L2/L3 documentation layers, creating AGENTS.md symlinks, or designing a fractal documentation system that helps AI quickly understand and…
Skill Claude CodeCodex
Review an existing code change, patch, pull request, commit, or working tree diff for defects, regressions, security issues, data risks, and important missing tests. Use when the primary task is evaluating code that already exists. Do not use for implementing an ordinary feature from scratch.
Skill Claude CodeCodex
Generate comprehensive project specifications for current project.
ViryaZheng/xmum-application-skills
Skill Claude CodeCodex
Use when preparing XMUM graduate application materials, official document requests, application timelines, language planning, recommendation logistics, and privacy-safe checklists.
Skill Claude CodeCodex
Interview one clear question before planning/building ideas, features, codebase maps, specs, UI flows, prototypes, or final plans.
Skill Codex
AoA/Abyss durable memory and owner orientation: use when ongoing work may depend on reviewed prior decisions, provenance, lifecycle/currentness, or an existing memo artifact, even when none is named. Also use to recall, review, or evolve a candidate, export, quarantine packet, object, corpus identity, lifecycle…
Skill Codex
A coordinator for the full user-interface design process, from choosing an approach through review and self-assessment. It connects these stages into one workflow and produces a shared report.
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: