Qarinah — Development skill for Claude Code
Your project remembers across coding agents.
How to install Qarinah
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open AjnasNB/qarinah and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Qarinah does
Your project remembers across coding agents. Compact, cited project memory for Codex, Claude Code, Cursor, and compatible tools.
Alternatives in Development
- Zeron — A native control plane for Claude Code, Codex, Cursor, Devin and other coding agents 1.4k ★
- Pi Skills — Skills for pi coding agent (compatible with Claude Code and Codex CLI) 1.3k ★
- Chops — 1,000+ macOS app to browse, edit, and manage skills across Claude Code, Cursor, Codex, Windsurf, and Amp 1.1k ★
README
Qarinah
Start a new coding-agent session without re-explaining your project.
Qarinah remembers the decisions, code relationships, tool outcomes, and Git worktree history your project is allowed to retain. A fresh Codex, Claude Code, Cursor, Kimi, Antigravity, Freebuff, CLI, or compatible MCP session receives only the relevant cited context instead of another full-history replay.
# Safe default: local project memory, metadata-only capture, no hosted account,
# no content disclosure, and no activation metrics.
npx qarinah@latest setup .
Successful setup now ends with the result to try next:
{
"capture": "metadata",
"firstRun": {
"message": "Qarinah is ready. It mapped your project and built a cited local memory graph.",
"tryNow": "npx qarinah query \"What decisions and outcomes should the next coding-agent session know?\" --format markdown",
"openGraph": "npx qarinah dashboard --serve"
}
}
Want to see the complete experience without touching your repository?
# Creates a populated isolated workspace in the operating-system temp directory.
# It does not configure your coding agents and does not enable telemetry.
npx qarinah demo
The demo returns its local dashboard path, a real retained decision, its event ID and SHA-256 evidence hash, plus the exact query that reconstructs the decision in a fresh terminal. [Follow the reproducible steps.](docs/GETTING-STARTED.md#try-the-isolated-demo-first)
[](https://qarinah.io/assets/qarinah-fresh-session-handoff.mp4)
The two-minute recording removes the fixture's temporary Session A transcript, opens a fresh Session B, and retrieves the
Related Skills
Project Memory: AgentSys
Cross-tool compatible: Works across all 5 platforms - Claude Code, OpenCode, Codex CLI, Cursor, and Kiro.
Hq Starter
Instructions you hand to your Claude Code or Codex agent to build your own HQ: a file-based memory system and
Byori
The multi-agent coding workspace that remembers your project — run Claude Code and Codex side by side over a s
Longhouse
Drive your actual coding agents, not another agent wrapper! Like codex/claude remote control but fully open so
Claude Reforge
Persistent memory and learning for Claude Code — remembers errors, solutions, and patterns across sessions
Coding Agent Session Search
Unified TUI and CLI to index and search your local coding agent session history across 11+ providers (Codex, C
Related Agents
Chadi Council
Dedicated decision council subagent. Four voices (Architect, Skeptic, Pragmatist, Critic) deliberate ambiguous
Sync Secretary
Grounded secretary for Sync. Use for source-cited questions about Claude and Codex council orchestration: task
Part 13: Memory Bridge (Give Coding Agents Your Brain)
When you spawn Codex or Claude Code to build something, they start blind. They don't know your architecture de