Skill Claude CodeCodex
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
not rated 2 8d ago A 21 tokens
copy · 94% MIT
Skill Claude CodeCodex
Use when implementing any feature or bugfix, before writing implementation code.
not rated 2 8d ago A 17 tokens
copy · 88% MIT
Skill Claude CodeCodex
Manage Git worktrees. Use when asked to create/switch/list/merge/remove worktrees, to keep multiple branches in parallel directories, or to clean up worktrees safely during development.
not rated 2 8d ago A 41 tokens
original MIT
Skill Claude CodeCodex
Use when you have a spec or requirements for a multi-step task, before touching code.
not rated 2 8d ago A 21 tokens
copy · 88% MIT
prove-it
29Skill Claude CodeCodex
A method for testing whether a proposed idea, design, or technical approach actually works through a falsifiable demonstration. Falsifiable means the test could show that the claim is wrong.
not rated 2 8d ago A 92 tokens
original MIT
Skill Claude CodeCodex
Load and use remote skills in the current conversation through bunx/npx skills use, from user-specified sources or a declared global/project catalog. Use when the user names a remote skill or catalog alias, asks to match a task against configured remote skills, or wants to view, add, edit, disable, or remove those…
not rated 2 8d ago A 79 tokens
original MIT
Skill Claude CodeCodex
Reassess a problem when a solution accumulates branches, special cases, or layers, or discussion repeatedly fails to converge. Use when the user asks to challenge the problem statement, validate the underlying requirement, or reframe a stuck design. Ordinary implementation difficulty alone is insufficient.
not rated 2 8d ago A 59 tokens
original MIT
Skill Claude CodeCodex
Rethink a project, product, workflow, or scenario around abundant AI agent execution capacity. Use when the user wants an AI-native or agent-native redesign, challenges limits inherited from earlier models or human-centered processes, or asks what unlimited agent and token budgets could enable for quality, automation…
not rated 2 8d ago A 79 tokens
original MIT
paseo-relay
33Skill Claude CodeCodex
Connect to other Paseo hosts through Relay using authorized pairing information; discover projects, workspaces and agents, inspect sessions and timelines, and perform explicitly requested remote operations. Use for cross-host Paseo access, including requests that call Relay “replay”.
not rated 2 8d ago A 56 tokens
original MIT
prune
34Skill Claude CodeCodex
Periodically slim down and restructure a project built largely by AI agents — remove dead code, duplicate implementations, abandoned approaches, unused features and APIs, over-defensive and over-abstracted code, and drifted docs, and fix the architectural causes that keep regrowing them — while preserving behavior…
not rated 2 8d ago A 118 tokens
original MIT
Skill Claude CodeCodex
A conversation mode that grows a visual map of a topic over time. The map stores ideas and their relationships locally, can continue across sessions, and lets users revisit the conversations behind its nodes.
not rated 2
changed 7d ago A 101 tokens
original MIT
MCP server Claude CodeCodexCursor +2
A Python MCP server for Qdrant memory storage and semantic search, compatible with Cursor-like MCP clients. Runs locally from the better-qdrant-mcp Python package.
not rated 1 11mo ago A
tokens not measured
original 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: