paper-writing-router

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute git commands (e.g., git rev-parse). This is used for repository discovery to determine the scope of the project. While legitimate, shell access remains a sensitive capability.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by reading external data from local files like memory/BRIEFING.md and paper/.agent/writing-contract.md to guide its routing logic. Without sanitization or boundary markers, malicious content within these files could potentially influence agent behavior.
  • Ingestion points: memory/BRIEFING.md, paper/.agent/writing-contract.md (processed via the Read tool).
  • Boundary markers: Absent. The skill does not use specific delimiters or instructions to ignore potential commands embedded in the data files.
  • Capability inventory: Bash (shell access), Read (file system access).
  • Sanitization: Absent. The content of the briefing files is used directly for task classification without validation or filtering.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 03:15 AM
Security Audit — agent-trust-hub — paper-writing-router