An autonomous AI agent CLI that investigates GitHub issues and generates comprehensive "starter briefs" for first-time open-source contributors — identifying likely root causes, relevant files, related past PRs, and difficulty ratings.
First-time open-source contributors often get stuck figuring out where to start. issuepilot is a true autonomous ReAct agent (not a hardcoded script). It fetches the issue, formulates search terms, inspects code files, checks past pull requests, evaluates repository guidelines, and dynamically decides when it has enough context to produce an actionable starter brief.
No installation required! Run instantly via npx:
npx issuepilot <url> On first run, issuepilot will launch an interactive wizard to prompt for your GITHUB_TOKEN and optional LLM keys.
Run the interactive setup wizard at any time:
npx issuepilot initissuepilot features a resilient multi-provider fallback engine with multi-key rotation for high availability:
| Provider | Env Key(s) | Free Tier Highlights |
|---|---|---|
| Pollinations.ai | (None required) | Zero key required, works out of the box! |
| Groq | GROQ_API_KEY |
30 RPM / 14,400 req/day ultra-fast Llama 3.3 — recommended free provider |
| GitHub Models | GITHUB_TOKEN |
Free LLM inference included with your GitHub PAT! |
| xAI / Grok |
GROK_API_KEY, XAI_API_KEY
|
Grok-beta & Grok-2 support |
| Cerebras | CEREBRAS_API_KEY |
30 RPM ultra-fast inference |
| OpenRouter | OPENROUTER_API_KEY |
Access to all OpenRouter free models (:free) |
| OpenAI / Custom |
OPENAI_API_KEY, LLM_BASE_URL
|
OpenAI or local Ollama / vLLM endpoints |
Supply multiple API keys for any provider (comma-separated or indexed) for automatic key rotation on rate limits:
GROQ_API_KEY_1="groq_key1"
GROQ_API_KEY_2="groq_key2"# Basic execution
npx issuepilot <issue-url>
# Output as raw JSON
npx issuepilot <issue-url> --json
# Save brief to a local file
npx issuepilot <issue-url> -o brief.md
# Copy brief directly to system clipboard
npx issuepilot <issue-url> -c
# Specify maximum agent investigation turns
npx issuepilot <issue-url> --max-turns 20
# Specify custom provider/model fallback sequence
npx issuepilot <issue-url> --fallback-models gemini/gemini-2.0-flash,groq/llama-3.3-70b-versatile,pollinations/openai Issue URL
│
▼
Fetch Issue & Guidelines
│
▼
┌──────────────────────────────────┐
│ ReAct AI Agent Loop │
│ (Forced Tool Choice Turn 0) │
└────────────────│─────────────────┘
│
┌─────────────────┼─────────────────┐
▼ ▼ ▼
search_codebase read_file search_history
(Qualifiers & (Line Slicing) (GET /search)
Path Ranking) │ │
│ └────────┬────────┘
└──────────────────────────┤
▼
submit_brief(...)
│
Schema Validator
│
Success
▼
Terminal Brief Output
Contributions are welcome! Please see CONTRIBUTING.md for details on setting up the local environment and running tests.
git clone https://github.com/ayesha-arbi/IssuePilot.git
cd IssuePilot
npm install
npm testMIT © IssuePilot Team