tailor-resume

Pass

Audited by Gen Agent Trust Hub on Nov 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: Analysis of the skill instructions and workflow revealed no malicious patterns, obfuscation, or unauthorized data access. The skill is designed to perform text transformation within the local environment.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes Job Descriptions, which are untrusted external inputs. 1. Ingestion points: Target JD (processed in SKILL.md). 2. Boundary markers: Not explicitly defined, but the skill relies on a local facts.json as the exclusive source of truth. 3. Capability inventory: Writes to the output/ directory (SKILL.md). 4. Sanitization: Step 5 implements a verification protocol requiring a mapping table for every bullet point to a source ID in facts.json, which prevents the agent from adopting instructions or fabrications from the untrusted JD input.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 7, 2026, 02:11 AM