osint-safe

name	osint-safe
description	Lawful, ethical open-source intelligence on public data only—no stalking, no bypassing access controls, no invasive targeting.
metadata	{"openclaw":{"requires":{"config":"[Truncated]"}}}

OSINT (safe, ethical)

When to use: The user wants to understand public information about a company, product, person’s professional footprint, or topic—using only lawful, consented, public sources.

Hard stops (refuse and explain)

• Doxxing, real-time physical tracking, or non-consensual intimate detail.
• Compromising accounts, phishing, or social engineering.
• Probing or exploiting systems without authorization.
• Collecting or inferring data about minors for non-safety purposes.

Allowed pattern

1. Scope — What decision does this support? (e.g. “Is this company legit for a B2B contract?”)
2. Surface — Company site, registered filings where public, official repos, press releases, mainstream news, public LinkedIn/job posts (high-level only).
3. Corroborate — Two independent public mentions when stakes matter.
4. Report — Facts vs inferences; unknowns; limits (“cannot verify private revenue”).
5. Safety — Do not facilitate harassment or coercion.

If only shady pathways would answer the question, say no and suggest legitimate alternatives (official channels, contracts, background vendors compliant with law).

Relationship to tools

• Prefer web_search / web_fetch within normal rate/respect boundaries.
• Do not instruct scraping behind logins or circumventing CAPTCHAs.