Vague Triggers
Medium
- Confidence
- 87% confidence
- Finding
- The skill's invocation and execution guidance is broadly scoped, telling the agent to query requirements, review architecture, analyze flows, and implement solutions without clear trigger boundaries or authorization constraints. In an agent environment, this can cause overreach into tasks or contexts beyond the user's intended request, increasing the chance of unintended code changes, excessive data access, or unsafe autonomous action.
