plan-interview

Pass

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill operates as a planning assistant using standard agent interactions. No unauthorized network activity or data exfiltration was detected.
  • [COMMAND_EXECUTION]: The skill includes an installation command using npx to add the skill from the author's repository. This is a standard and expected deployment step.
  • [PROMPT_INJECTION]: The skill ingests data from external codebase files (README.md, AGENTS.md) during exploration and uses reference templates. Ingestion points: Phase 2 and Phase 3. Boundary markers: Absent. Capability inventory: AskUserQuestion tool usage, file system access, and automated implementation start. Sanitization: Absent. This defines an indirect prompt injection surface, which is mitigated by the requirement for user approval of the implementation plan.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 22, 2026, 12:14 AM