ralplan
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted user input via the 'task description' parameter. This input is processed by a multi-agent pipeline (Planner, Architect, Critic) and used to generate plans that drive execution-capable skills like 'team' and 'ralph'.\n- Evidence Chain:\n
- Ingestion points: 'task description' argument in the '/oh-my-claudecode:ralplan' command (SKILL.md).\n
- Boundary markers: The skill does not define explicit delimiters or 'ignore' instructions when passing the task description to the planning agents.\n
- Capability inventory: The workflow results in calls to 'oh-my-claudecode:team' and 'oh-my-claudecode:ralph' (SKILL.md, steps 7 and 8), which are designed for code execution and implementation.\n
- Sanitization: While a 'Pre-Execution Gate' ensures prompts are not too vague, it lacks mechanisms to sanitize or validate the input against malicious injection patterns.
Audit Metadata