implementation-planning-v3
Pass
Audited by Gen Agent Trust Hub on Jul 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to identify and run shell commands for "minimum checks" (e.g., lint, typecheck, compile, and basic tests) as well as "baseline commands" to verify the project state. These commands are derived from the technical design or project context and are executed locally. There is no explicit sanitization or validation of these command strings before execution.
- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it relies on external technical design documents to generate task contracts.
- Ingestion points: The agent reads confirmed technical designs, requirement summaries, and review reports from the local workspace (specified in
SKILL.md, Step 2). - Boundary markers: The instructions do not define clear delimiters or "ignore embedded instructions" warnings when processing the technical design content.
- Capability inventory: The skill allows the agent to write files (implementation task documents) and execute shell commands based on the ingested content.
- Sanitization: There is no evidence of filtering or escaping logic to prevent malicious instructions in the design documents from overriding the agent's behavior.
Audit Metadata