dark-social-attributor
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
- [SAFE]: The skill consists entirely of instructional Markdown and reference documentation. No code, scripts, or automated execution logic are provided.
- [CREDENTIALS_UNSAFE]: The skill advocates for secure development practices by explicitly stating that API keys should never be stored within the skill or source control (evidence in references/connectors.md).
- [INDIRECT_PROMPT_INJECTION]: The skill instructions include a specific security guardrail for handling untrusted data. It directs the agent to treat scraped content and third-party exports as untrusted and forbids following any instructions embedded within that evidence (evidence in SKILL.md).
Audit Metadata