Prospector

This skill should be used when the user wants to find leads, prospects, or contacts matching their ICP (Ideal Customer Profile). It searches for companies via Exa and enriches contacts via Apollo, outputting to CSV and optionally syncing to Attio CRM. MANDATORY TRIGGERS: "find leads", "prospecting", "ICP search", "find contacts", "lead generation", "/prospector"

What to consider before installing: 1) Metadata mismatch — the skill's registry metadata lists no required env vars, but the skill needs Exa and Apollo API keys (and optionally Attio). Confirm you understand and accept providing those keys. 2) Inspect the code yourself (scripts/prospector.py) before running: it makes HTTP calls to api.exa.ai, api.apollo.io, and api.attio.com and writes results to ~/Desktop and ~/.config/prospector/config.json. 3) Prefer environment variables over saving keys to disk; if you must save them, verify permissions (the script sets chmod 600). 4) Be cautious about allowing the setup flow to append exports to your shell profile — that modifies files that may be synced/backed up. 5) If you have concerns, run the skill in an isolated environment or VM, or remove the parts that write to your shell profile and config file. 6) If you proceed, ensure API keys have least privilege and can be rotated/revoked easily.

Type: OpenClaw Skill Name: prospector Version: 1.0.0 The OpenClaw skill 'prospector' is benign. It performs lead generation by interacting with Exa, Apollo, and optionally Attio APIs, saving results to a CSV on the user's Desktop. API keys are handled securely by either reading environment variables or storing them in `~/.config/prospector/config.json` with strict `0o600` file permissions, as explicitly enforced in `scripts/prospector.py` and `commands/setup.md`. All network requests are directed to the legitimate API endpoints of the stated services. The instructions in `SKILL.md` and `commands/setup.md` are clear, directly align with the skill's purpose, and contain no evidence of prompt injection attempts or instructions for unauthorized actions.