#361 · Primary category: AI Coding Assistants
claude-smart
Turns corrections into Preferences, Project-specific skills, and Shared skills for Claude Code, Codex, and OpenCode.
Project last updated:08/28/26
GitHub Stars
774
Forks
87
Contributors
9
License
Apache-2.0
Why we included this project
Corrections you make to a coding assistant rarely stick, so you end up re-explaining the same thing in every session. claude-smart is a plugin for Claude Code, Codex, and OpenCode that captures those corrections, along with command sequences that already work, and turns them into durable rules injected back into the system prompt on each new session. Instead of replaying old transcripts, it distills your guidance into compact, deduplicated rules like 'always pass --run to npm test in this repo.' Everything runs locally with an in-process embedder and no external API key, and the rules maintain themselves by merging, superseding, and archiving as your project evolves. The two-tier profile-and-playbook design is easy to inspect through the bundled dashboard, which matters if you want the assistant to change behavior over time rather than just recall past context.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
superpowers
An agentic skills framework & software development methodology that works.
ECC
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
skills
Skills for Real Engineers. Straight from my .agents directory.
opencode
The open source coding agent.
andrej-karpathy-skills
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.