Automatically preserves context for Claude Code, preventing data loss from compaction. Ideal for developers and operations teams using Claude AI. Integrates with 7 hooks, 187 semantic patterns, and 3 MCP tools. Works out of the box with zero configuration.
git clone https://github.com/Capnjbrown/c0ntextKeeper.gitc0ntextKeeper automatically captures and archives your Claude Code conversations before compaction occurs, eliminating the need to rediscover solutions across sessions. It uses 7 hooks integrated into Claude Code and 187 semantic patterns to intelligently extract problems, solutions, and architectural decisions from your work. The system provides searchable access via CLI commands and MCP tools, with automatic redaction of API keys and credentials for security. Setup requires a single command with zero configuration needed, and it operates with <10ms performance impact. Developers and operations teams can retrieve past context naturally by asking Claude questions about previous work.
[{"step":"Trigger c0ntextKeeper before starting a critical task. For example, run `c0ntextKeeper --project api-gateway-rewrite --goal 'implement rate limiting'` in your terminal or invoke it via your IDE's AI assistant.","tip":"Use the `--verbose` flag to see detailed logs of preserved context. This helps debug if something is missing."},{"step":"Verify the preserved context by reviewing the summary output. Check that all [LIST_KEY_CONTEXT_ITEMS] are included and no uncommitted changes are missed.","tip":"If you're working in a team, share the context summary with collaborators to ensure alignment on the current state."},{"step":"Proceed with your task (e.g., refactoring, debugging, or deploying). c0ntextKeeper will silently monitor for context changes and alert you if critical data is lost.","tip":"For long-running tasks, periodically run `c0ntextKeeper --check` to ensure context integrity."},{"step":"If context loss is detected, use the preserved state to restore files or dependencies. For example, run `c0ntextKeeper --restore --commit a1b2c3d` to revert to a known good state.","tip":"Combine with `git` or your version control system to create a backup tag (e.g., `git tag context-save-2024-05-20`) for easy rollback."},{"step":"After completing the task, run `c0ntextKeeper --cleanup` to remove temporary context files and free up disk space.","tip":"Set up a cron job or GitHub Action to run `c0ntextKeeper --cleanup` weekly to avoid accumulating stale context data."}]
Retrieve past authentication implementations without manual searching
Access architectural decisions and design reasoning from previous sessions
Find solutions to recurring errors without re-debugging from scratch
Onboard teammates by sharing searchable context archives
No install command available. Check the GitHub repository for manual installation instructions.
git clone https://github.com/Capnjbrown/c0ntextKeeperCopy 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.
Use c0ntextKeeper to preserve the full context of [PROJECT_NAME] for [TASK_OR_GOAL]. Ensure all relevant files, dependencies, and recent changes are captured before proceeding with [SPECIFIC_ACTION]. Include the following critical details: [LIST_KEY_CONTEXT_ITEMS]. Verify the context is intact by summarizing the preserved state.
### Context Preservation Report for Project: `api-gateway-rewrite` **Preserved Context Summary:** - **Repository State:** 42 files modified (12 new, 8 deleted, 22 updated) in the last 7 days. Key changes include: - `/src/routes/v2/auth.py` (new): Added JWT validation middleware - `/config/k8s/deployment.yaml` (updated): Increased replica count to 5 - `/tests/integration/test_auth.py` (new): Added 15 test cases for token refresh - **Dependencies:** Updated `requirements.txt` with `fastapi==0.109.0` and `pydantic>=2.5.0`. Lock file shows no conflicts. - **Recent Commits:** 3 commits by `dev-team` focusing on auth improvements (SHA: `a1b2c3d`, `e4f5g6h`, `i7j8k9l`). - **Open Issues:** #42 (priority: high) - "Handle rate limiting for auth endpoints" (assigned to @alice-dev). - **Environment:** Docker image `api-gateway:2.1.0` built successfully. Kubernetes cluster `prod-cluster-1` is running v1.27.3. **Verification:** ✅ All 42 modified files accounted for. No uncommitted changes detected. Dependencies resolved. Critical paths (`/src/auth/*`) preserved with full history. **Next Steps:** 1. Proceed with deploying the updated auth middleware to staging. 2. Monitor error logs for `JWTValidationError` in the next 2 hours. 3. If issues arise, roll back to commit `a1b2c3d` using `git reset --hard a1b2c3d`. *Context preserved at: 2024-05-20T14:32:15Z*
Automate your browser workflows effortlessly
AI assistant built for thoughtful, nuanced conversation
Agents that listen, think and act for you.
Get more done every day with Microsoft Teams – powered by AI
Agentic AI Workflow platform
Connected workspace for docs, wikis, and projects
Take a free 3-minute scan and get personalized AI skill recommendations.
Take free scan