Identity Trace

Use the bundled standard-library CLI as the deterministic collection and reporting layer:

python3 <skill>/scripts/identity_trace.py run <identifier>

Do not install tools automatically. Let the CLI detect optional local tools and credentials, run every applicable available source by default, and record unavailable or failed sources in the case bundle.

Workflow

1. Run doctor --json when source availability is unknown. Treat credential presence as sensitive; report names and availability, never values.
2. Run run with the known identifiers. Add typed flags when combining an email, phone, username, name, or birth date. Read references/cli-reference.md for exact options.
3. Read the emitted queries.json. Use available web or browser search tools for useful pivots, especially full-name and birth-date searches that the script cannot execute itself.
4. Record selected search results with ingest. Include the exact query, URL, title, snippet, observation time, and related input IDs. Do not ingest a result merely because it resembles the target.
5. Run validate after ingestion or manual transfer of a case bundle.
6. Return the bundle path, source coverage, material observations, and collection failures. Keep hypotheses separate from the evidence report.