dynamo-recipe-runner

Pass

Audited by Gen Agent Trust Hub on Jun 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes kubectl to interact with Kubernetes clusters, performing actions such as applying manifests, checking resource status, and managing port-forwarding sessions.
  • [COMMAND_EXECUTION]: It executes a local Python script, scripts/recipe_tool.py, for the purpose of discovering and validating available deployment recipes.
  • [PROMPT_INJECTION]: The skill ingests untrusted configuration data from files in the recipes/ tree (Ingestion points). No specific boundary markers or instructions to disregard embedded commands are used when this data is processed by the agent (Boundary markers). The skill is granted significant capabilities, including shell command execution and kubectl cluster mutations (Capability inventory). To mitigate potential risks, the skill uses scripts/recipe_tool.py to validate manifest files and identify configuration errors or placeholders before any cluster-mutating actions are taken (Sanitization).
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 12, 2026, 07:42 PM
Security Audit — agent-trust-hub — dynamo-recipe-runner