OpenContext

🥈Silver

A personal context store for AI agents and assistants. reuse your existing coding agent CLI (Codex/Claude/OpenCode) with built‑in Skills/tools and a desktop GUI to capture, search, and reuse project knowledge across agents and repos.

484220Updated 2w ago

Intermediate30min to implementautomation

Saves ~120 min per use

Quick InstallView Source

git clone https://github.com/0xranx/OpenContext.git

Works with:

Claude

Overview

About This Skill

https://0xranx.github.io/OpenContext/en/usage/

How to Use

[{"step":"Install OpenContext CLI and desktop GUI","action":"Run `npm install -g @opencontext/cli` and launch the desktop app. Authenticate with your Git provider (GitHub/GitLab) to sync repos.","tip":"Use `oc --version` to verify installation. Ensure your Git credentials have read access to the target repos."},{"step":"Index your project knowledge","action":"Run `oc index [REPO_PATH]` or use the GUI to select a repo. Wait for indexing to complete (progress shown in CLI/GUI).","tip":"For large repos, indexing may take 5-10 minutes. Use `oc index --depth 2` to limit depth for faster results."},{"step":"Query with context-aware prompts","action":"Use the CLI (`oc query \"[YOUR_PROMPT]\"`) or GUI search bar to ask questions like 'Show me all API endpoints that handle payments.'","tip":"Prefix queries with repo names (e.g., `oc query \"payment-gateway: show me error handling patterns\") for focused results."},{"step":"Reuse and refine knowledge","action":"Save outputs to OpenContext's memory with `oc save \"[SUMMARY]\" --tags [TAGS]`. Reference saved contexts in future queries (e.g., 'Based on my saved context \"error-handling-patterns\", update the retry logic.').","tip":"Use tags like `bug-fix`, `refactor`, or `documentation` to organize saved contexts for quick retrieval."},{"step":"Automate with agents","action":"Configure your coding agent (e.g., Claude Code) to use OpenContext as a tool. Add `\"tools\": [\"opencontext\"]` to your agent's config file.","tip":"Test agent integration with `oc agent-test \"[QUERY]\"` to ensure the agent can access OpenContext's indexed data."}]

Use Cases

Automatically load project context before starting a coding session with an AI agent.

Create and manage a global library of project knowledge accessible across different repositories.

Search for relevant documentation quickly to assist in coding tasks.

Persist learned information from AI interactions to improve future responses.

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/0xranx/OpenContext

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Use OpenContext to [ACTION] for [PROJECT/REPO]. Start by [INITIAL_STEP] and reference past work where relevant. For example: 'Use OpenContext to review the error handling patterns in the [PROJECT_NAME] repo. Start by searching for 'try-catch' blocks in the last 6 months of commits and summarize common anti-patterns.'

Example Output

### OpenContext Analysis: Error Handling Patterns in 'payment-gateway' Repo

**Scope**: Reviewed 47 commits from the last 6 months (Jan 2024 - Jun 2024) across 12 files in the `payment-gateway` repo. Focused on `try-catch` blocks, `Promise` handling, and error logging.

**Key Findings**:
1. **Inconsistent Error Logging**: 68% of error handlers use `console.error()` without structured logging (e.g., no correlation IDs or stack traces). Only 3 files in `src/services/` use `winston` or `pino`.
2. **Overly Broad Catch Blocks**: 42% of `try-catch` blocks catch `Error` instead of specific exceptions (e.g., `PaymentProcessingError`). This masks critical failures like network timeouts.
3. **Missing Retry Logic**: 5 files lack retry mechanisms for transient errors (e.g., database timeouts). Only `src/integrations/stripe.js` implements exponential backoff.
4. **Uncaught Promise Rejections**: 18% of async functions lack `.catch()` handlers, risking silent failures in production.

**Recommended Actions**:
- **Immediate**: Add structured logging to `src/utils/errorHandler.js` (template provided in `docs/error-handling.md`).
- **Short-term**: Refactor 8 files to catch specific exceptions (e.g., `StripeError`, `DatabaseTimeoutError`).
- **Long-term**: Implement retry logic in `src/services/paymentService.js` using the `p-retry` library (see `examples/retry-patterns.js`).

**References**:
- Past PR #124 (Mar 2024) introduced `winston` but was only partially adopted.
- Issue #201 tracks silent failures in the `/checkout` endpoint.

**Next Steps**: OpenContext has tagged all affected files with `error-handling` for easy tracking. Would you like me to generate a PR draft with these changes?