sailpoint-toxic-access-detector

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes external data from the SailPoint API (e.g., identity names, source names, and policy descriptions) to generate audit reports. This is a common and expected characteristic for security auditing skills. \n
  • Ingestion points: identity search results and SoD policy definitions retrieved via sail api commands in SKILL.md. \n
  • Boundary markers: Absent; fetched data is used directly in report generation. \n
  • Capability inventory: Shell execution of the sail CLI for data retrieval and environment management in SKILL.md. \n
  • Sanitization: Absent. \n- [COMMAND_EXECUTION]: The skill uses placeholders such as {name}, {id}, and {accessId} within shell commands for the sail CLI. These allow the agent to interact dynamically with the environment based on user input or API responses, which is the intended functionality of the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 11:28 AM