cover-letter-builder

Pass

Audited by Gen Agent Trust Hub on May 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands (pdftotext, sed, pandoc) for text extraction from PDF and LaTeX files. While these are standard utilities, executing them on user-provided file paths carries a risk of command injection if the underlying platform does not sanitize arguments.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8) because it processes untrusted content from external sources to generate output.
  • Ingestion points: The skill reads job descriptions and CV/Resume files (PDF, Markdown, LaTeX, TXT) provided by the user or discovered during a recursive repository scan in Step 1.
  • Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the ingested text as data rather than instructions, which could allow malicious text within a job posting to manipulate the agent's behavior.
  • Capability inventory: The agent has the capability to read files, execute shell commands for processing, and write new Markdown files to the local filesystem (Step 8).
  • Sanitization: No explicit sanitization or validation of the ingested text or file paths is defined in the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
May 7, 2026, 11:21 AM
Security Audit — agent-trust-hub — cover-letter-builder