agentation-watch-mode

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it continuously ingests and processes external data from the agentation_agentation_watch_annotations tool.
  • Ingestion points: The agentation_agentation_watch_annotations tool in SKILL.md fetches user-provided annotations from an external browser session.
  • Boundary markers: The skill lacks explicit instructions or delimiters to isolate the annotation content from the agent's core logic, increasing the risk that the agent will follow instructions embedded in the data.
  • Capability inventory: The agent is directed to 'Make the requested fix' (SKILL.md) based on these annotations, which likely involves file system modifications or code execution capabilities.
  • Sanitization: There is no evidence of validation or filtering for the content of the annotations before they are used to determine the agent's actions.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 05:50 PM