retention-system

Pass

Audited by Gen Agent Trust Hub on Mar 8, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [SAFE]: No malicious prompt injection, obfuscation, or data exfiltration patterns were detected in the skill instructions.
  • [NO_CODE]: The skill does not contain or execute any scripts, binaries, or external code dependencies.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its research methodology.
  • Ingestion points: The skill uses WebSearch and WebFetch tools to gather industry benchmarks and methodologies (SKILL.md).
  • Boundary markers: No specific delimiters or instructions to ignore embedded commands are present in the research guidelines.
  • Capability inventory: The skill possesses file manipulation (Read, Write, Edit, Grep, Glob) and network access (WebSearch, WebFetch) capabilities.
  • Sanitization: There is no evidence of sanitization or validation of the data retrieved from external web sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 8, 2026, 04:35 PM