recruit-score

Pass

Audited by Gen Agent Trust Hub on Aug 3, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted external data (candidate resumes, LinkedIn URLs, and interview notes) which creates a surface for indirect prompt injection. A candidate could potentially embed instructions within their resume or LinkedIn profile designed to override the scoring rubric or force a 'STRONG HIRE' recommendation.
  • Ingestion points: The <candidate> argument in the /recruit score command, which is intended to receive resume text, external profile URLs, and interview notes.
  • Boundary markers: Absent. There are no instructions to wrap external content in delimiters or to ignore instructions embedded within the processed data.
  • Capability inventory: The skill's primary capability is generating a text-based decision memo saved to a local markdown file (RECRUIT-SCORE-[Candidate].md). It does not appear to have access to high-risk tools like network exfiltration or arbitrary shell execution.
  • Sanitization: Absent. The skill does not provide instructions for the agent to sanitize, escape, or validate the inputs provided in the candidate data.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 3, 2026, 06:04 PM
Security Audit — agent-trust-hub — recruit-score