What problem does it solve? OSINT investigations often start without a clear scope or an approved plan, leading to unfocused research and unverifiable findings. This Skill structures the first two phases of a Spotlight investigation: a user-approved brief and a validated, tool-grounded methodology before any research begins. ## Core Features & Use Cases - Brief Conversation with Approval Gate: Scrapes lead URLs, asks clarifying questions, and persists the approved direction to brief-directions.txt. - Investigator Planning Spawn: Launches an investigator agent in PLANNING mode that writes methodology.json with tool selections from either the Navigator CLI or a local osint-tools index. - Methodology Validation and Correction: Runs validate-methodology-navigator.py, re-spawns the investigator with bounded correction prompts on failure, and records an attributable approval receipt. - Optional RLM Proposal: On frontier setups, proposes deterministic or hybrid Gemma4 RLM analysis with an explicit opt-in decision recorded in methodology.json. - Use Case: A journalist receives a lead about a Swiss foundation; the Skill scrapes the source URL, confirms scope with the user, spawns a planning agent that selects registry lookup tools, validates the methodology, and gates research until the user approves. ## Quick Start Start a new Spotlight investigation from my lead and walk me through approving the brief and methodology before any research begins.
