ln-64-community-announcer

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data such as repository documentation, source code diffs, and commit messages, which could theoretically contain malicious instructions.
  • Ingestion points: The skill reads repository communication guidance, authoritative documentation, release notes, and source code diffs from the target GitHub repository (SKILL.md).
  • Boundary markers: The skill implements a robust Human-in-the-loop (HITL) mechanism, requiring the agent to "Wait for explicit approval" and "Present the exact title and full Markdown body to the user before creating external state" (SKILL.md).
  • Capability inventory: The skill possesses GitHub API write access to create discussions and file system access to write temporary draft files (SKILL.md).
  • Sanitization: The instructions explicitly require "Safely escaped API input" and writing drafts to temporary files to "avoid shell interpolation errors" (SKILL.md).
  • [COMMAND_EXECUTION]: The skill utilizes the GitHub CLI (gh) and git to perform its primary functions.
  • The usage is restricted to authenticated repository operations and is explicitly gated by user approval of the generated content (SKILL.md).
  • [DATA_EXPOSURE]: The skill includes explicit instructions to protect sensitive information.
  • It mandates: "Do not expose repository tokens, category node IDs, or other credentials in the announcement" (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 05:02 AM
Security Audit — agent-trust-hub — ln-64-community-announcer