# Spotlight

> Runtime-agnostic OSINT investigation system for journalists. Spotlight gives an agent runtime a reporting methodology, an investigator and fact-checker loop, human review gates, and a reusable case archive.

## Start here

- [Install](https://spotlight.buriedsignals.com/setup.html): public setup flow for journalists
- [Landing page](https://spotlight.buriedsignals.com/): public product overview
- [Documentation index](https://spotlight.buriedsignals.com/docs/README.md): operator manual and reading order
- [Runtime contract](https://spotlight.buriedsignals.com/AGENTS.md): machine-readable verb registry, agent manifests, skill registry, and case structure
- [Full public AI reference](https://spotlight.buriedsignals.com/llms_full.txt): flattened summary for one-shot ingestion
- [GitHub repo](https://github.com/buriedsignals/spotlight): source, issues, releases

## What Spotlight does

- [README](https://spotlight.buriedsignals.com/README.md): overview, supported runtimes, dependencies, and public documentation map
- [Investigation pipeline](https://spotlight.buriedsignals.com/docs/investigating.md): brief, methodology, cycles, Gate 1, ingestion, and readiness criteria
- [Fact-checking model](https://spotlight.buriedsignals.com/docs/fact-checking.md): independent verification pass, SIFT workflow, verdict taxonomy, and evidence trails
- [Monitoring](https://spotlight.buriedsignals.com/docs/monitoring.md): how ongoing monitoring feeds new leads back into investigations
- [Integrations](https://spotlight.buriedsignals.com/docs/integrations.md): dev-browser, Junkipedia, OSINT Navigator, Scoutpost, and manifest/preflight behavior
- [Runtime wiring](https://spotlight.buriedsignals.com/docs/runtimes.md): pi, Hermes, Goose, Codex, Gemini, and local OpenAI-compatible setups

## Core concepts

- **Pipeline:** Preflight -> Brief -> Methodology -> Execution cycles -> Gate 1 -> Ingestion
- **Two-agent model:** investigator executes the research; fact-checker independently verifies the claims
- **Editorial control:** the user approves the brief, methodology, Gate 1 summary, and ingestion
- **Readiness criteria:** minimum findings, source independence, no unresolved disputes, affected perspective, document trail, and gap assessment
- **Evidence grounding:** scrape before cite, keep local copies, and archive sources before relying on them
- **Sensitive mode:** strips web search and scraping so investigations can run local-only

## Public docs

- [Docs home](https://spotlight.buriedsignals.com/docs/README.md)
- [Structure](https://spotlight.buriedsignals.com/docs/structure.md)
- [Runtimes](https://spotlight.buriedsignals.com/docs/runtimes.md)
- [Integrations](https://spotlight.buriedsignals.com/docs/integrations.md)
- [Investigating](https://spotlight.buriedsignals.com/docs/investigating.md)
- [Fact-checking](https://spotlight.buriedsignals.com/docs/fact-checking.md)
- [Monitoring](https://spotlight.buriedsignals.com/docs/monitoring.md)
- [Recovery](https://spotlight.buriedsignals.com/docs/recovery.md)

## Skills

- [spotlight](https://spotlight.buriedsignals.com/skills/spotlight/SKILL.md): top-level investigation orchestrator
- [review](https://spotlight.buriedsignals.com/skills/review/SKILL.md): post-Gate-1 review artifact and feedback loop
- [integrations](https://spotlight.buriedsignals.com/skills/integrations/SKILL.md): routing layer for external tools
- [ingest](https://spotlight.buriedsignals.com/skills/ingest/SKILL.md): vault ingestion and registry updates
- [monitoring](https://spotlight.buriedsignals.com/skills/monitoring/SKILL.md): monitoring orchestration across runtimes
- [web-archiving](https://spotlight.buriedsignals.com/skills/web-archiving/SKILL.md): archive workflow and chain of custody
- [content-access](https://spotlight.buriedsignals.com/skills/content-access/SKILL.md): paywall and access-method handling
- [osint](https://spotlight.buriedsignals.com/skills/osint/SKILL.md): OSINT tool routing and catalog
- [investigate](https://spotlight.buriedsignals.com/skills/investigate/SKILL.md): investigation techniques and reference playbooks
- [follow-the-money](https://spotlight.buriedsignals.com/skills/follow-the-money/SKILL.md): financial and ownership tracing methods
- [social-media-intelligence](https://spotlight.buriedsignals.com/skills/social-media-intelligence/SKILL.md): account authenticity, coordination, and narrative tracking

## Agents and schemas

- [investigator prompt](https://spotlight.buriedsignals.com/agents/investigator.md)
- [fact-checker prompt](https://spotlight.buriedsignals.com/agents/fact-checker.md)
- [findings schema](https://spotlight.buriedsignals.com/schemas/findings.schema.json)
- [fact-check schema](https://spotlight.buriedsignals.com/schemas/fact-check.schema.json)
- [methodology schema](https://spotlight.buriedsignals.com/schemas/methodology.schema.json)
- [investigation log schema](https://spotlight.buriedsignals.com/schemas/investigation-log.schema.json)
- [summary schema](https://spotlight.buriedsignals.com/schemas/summary.schema.json)

## Runtime contract

- The fixed verb registry is: `fetch`, `search`, `read-file`, `write-file`, `edit-file`, `list-files`, `grep-files`, `execute-shell`, `spawn-agent`, `wait-agent`, `invoke-skill`, `query-vault`, `vault-write`.
- `fetch` and `search` are backed by `firecrawl`.
- `query-vault` is backed by `BUN_INSTALL="" qmd query`.
- `vault-write` is backed by the `obsidian` CLI.

## Compatibility

- Canonical full reference: https://spotlight.buriedsignals.com/llms_full.txt