social-check

Pass

Audited by Gen Agent Trust Hub on Jul 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a bundled linter script (scripts/lint-post.mjs) via the bun or node runtime to perform deterministic checks on social media drafts, such as character counts and hashtag placement.\n- [PROMPT_INJECTION]: The skill processes untrusted user-supplied social media drafts to provide validation and stylistic judgment, which presents a surface for indirect prompt injection.\n
  • Ingestion points: User-provided draft text ingested through SKILL.md workflows.\n
  • Boundary markers: None; the agent processes the draft content directly for judgment.\n
  • Capability inventory: Execution of a bundled script and reading local context files (social-context.md).\n
  • Sanitization: The linter script performs mechanical validation (regex-based) but does not sanitize the draft content for malicious instructions before agent evaluation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 19, 2026, 09:01 PM
Security Audit — agent-trust-hub — social-check