Built for Claude Code but works in any agent that supports skills, easily recall what you discussed with your favorite coding agents, what decisions you made, and where you left off so you can pick up where you left off as easily as asking "where did we leave off on XYZ..."
git clone https://github.com/alexknowshtml/kuato.gitBuilt for Claude Code but works in any agent that supports skills, easily recall what you discussed with your favorite coding agents, what decisions you made, and where you left off so you can pick up where you left off as easily as asking "where did we leave off on XYZ..."
1. **Start a session:** Begin your coding project with Claude Code or another agent, and explicitly ask it to use the `kuato` skill for context tracking. Example: 'Let's use kuato to track our progress on the API refactor.' 2. **Pause naturally:** When you need to stop, say something like 'Let's pause here and save our progress with kuato.' The agent will summarize the session. 3. **Resume later:** Later, ask 'Where did we leave off on [PROJECT_NAME]?' or 'What were the key decisions for [FEATURE_NAME]?' to recall the context. 4. **Refine the summary:** If the summary is too broad, ask follow-ups like 'What specific code changes did we make to the auth middleware?' to drill down. 5. **Integrate into workflow:** Use `kuato` at the end of each coding session to document progress, then review the summary before your next session to quickly regain context. For teams, share the `kuato` summaries in your project’s Slack channel or GitHub issue tracker.
No install command available. Check the GitHub repository for manual installation instructions.
git clone https://github.com/alexknowshtml/kuatoCopy the install command above and run it in your terminal.
Launch Claude Code, Cursor, or your preferred AI coding agent.
Use the prompt template or examples below to test the skill.
Adapt the skill to your specific use case and workflow.
Recall our previous discussion about [PROJECT_NAME] in our coding session. Summarize the key decisions we made, the code changes implemented, and where we left off. Focus on actionable next steps and unresolved issues. Example: 'Recall our work on the user authentication system. What did we decide about the password hashing approach, and where did we pause the implementation?'
Here’s a summary of our last session on the **e-commerce checkout flow** (Project: `storefront-v2`), which we paused on **Friday at 3:45 PM** after implementing the cart persistence feature:
**Key Decisions:**
1. **Database Schema:** We finalized the `carts` table with columns for `user_id`, `session_id`, `items` (JSONB), and `expires_at`. The `items` field will store an array of `{product_id, quantity, price_at_time}` to preserve historical pricing.
2. **Security:** We agreed to use JWT for session management but deferred the refresh token implementation. Instead, we’ll rely on short-lived (15-minute) access tokens for now.
3. **Frontend Integration:** The React component `CartDrawer` now syncs with the backend via `/api/cart`, but we need to handle race conditions when multiple tabs are open.
**Where We Left Off:**
- The `POST /api/cart/items` endpoint works, but the `PATCH /api/cart/items/:product_id` endpoint throws a 500 error when updating quantities. The error suggests a race condition in the `UPDATE` query.
- We need to add a loading state to the `CartDrawer` to prevent duplicate submissions.
- The `DELETE /api/cart` endpoint is stubbed but untested.
**Next Steps:**
1. Debug the `PATCH` endpoint by adding a `console.log` in the `cartService.updateItem()` function to inspect the query parameters.
2. Write a test case for the `DELETE` endpoint using `vitest` (we’re mocking the database with `pg-mock`).
3. Schedule a follow-up to discuss refresh tokens—should we use Redis for session storage, or stick with the JWT approach?
**Files Modified:**
- `backend/services/cartService.js` (added `updateItem` and `deleteCart` stubs)
- `frontend/src/components/CartDrawer.jsx` (added loading state)
Would you like me to draft the fix for the `PATCH` endpoint or prioritize the `DELETE` tests?AI assistant built for thoughtful, nuanced conversation
Modern scheduling made simple for teams
IronCalc is a spreadsheet engine and ecosystem
Customer feedback management made simple
Enterprise workflow automation and service management platform
Automate your spreadsheet tasks with AI power
Take a free 3-minute scan and get personalized AI skill recommendations.
Take free scan