retention-verification

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform operations using shell commands including 'ae-cli' and 'cat', and to write data to temporary files in '/tmp/'. While standard for this skill's function, it utilizes direct shell interaction.
  • [INDIRECT_PROMPT_INJECTION]: The skill interpolates untrusted user data such as report IDs, project IDs, and event names into shell commands and JSON configuration templates.
  • Ingestion points: User-provided inputs for parameters such as 'initial_event', 'project_id', and 'report_id' are utilized in 'SKILL.md'.
  • Boundary markers: The instructions do not define clear delimiters or provide guidance to the agent to treat these inputs as data rather than instructions, potentially leading to command or JSON injection.
  • Capability inventory: The skill has the capability to execute commands via the 'ae-cli' and perform file system writes to temporary directories.
  • Sanitization: There are no explicit validation or sanitization routines mentioned to filter or escape the interpolated user data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 07:21 AM
Security Audit — agent-trust-hub — retention-verification