My personal collection of specialized skills for AI agents, organized as discrete, reusable capabilities.
Skill Claude CodeCodex
Make long autonomous work survive context-window wipes and heavy compactions by externalizing agent state to disk (PROMPT.md / PLAN.md / HISTORY.md). Use proactively when starting a multi-step task likely to span sessions, resets, or budget exhaustion: long builds, research harnesses, multi-commit features…
Skill Claude CodeCodex
Create atomic, logically-grouped git commits from uncommitted changes. Analyzes files, buckets by functionality, and commits in dependency order. Use when user has many changes to commit, wants to "break down" changes, or asks for commit strategy. Examples: user: "commit all these changes" → analyze, group, and commit…
Skill Claude CodeCodex
Write well-formatted changelogs following Keep a Changelog and SemVer, from scratch or from existing records (git history, tags, release notes, an existing CHANGELOG.md). Buckets changes into Added/Changed/Deprecated/Removed/Fixed/Security, derives the next version with a semver bump, and preserves the file's existing…
Skill Claude CodeCodex
Use this skill whenever an agent needs to commit a precise, character-level subset of file changes to a Git repository without disturbing the working directory. Triggers include: staging only part of a file's edits, committing one logical change while leaving other in-progress work untouched on disk, or when…
Skill Claude CodeCodex
Design software systems using the state machine (statechart) pattern. Use this skill whenever a user wants to model application logic, UI flows, async operations, protocols, workflows, or business processes as a state machine. Triggers include: "design a state machine", "model this as states", "create a statechart"…
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: