paidf-orchestration-write-dag
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions to execute shell commands using
kubectl,curl, andmaketo manage the deployment lifecycle and perform readiness checks on the Kubernetes cluster. These operations are intended for infrastructure management and are gated by explicit user confirmation for state-changing actions. - [DYNAMIC_EXECUTION]: The skill generates Python source code for Airflow DAGs and callables at runtime based on user input. It verifies the integrity of these generated files by attempting to import them as modules using the Python interpreter, which is consistent with the skill's primary purpose.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes user-provided pipeline descriptions and reads external documents like
launchable.mdto inform code generation. Ingestion points: User-supplied pipeline stage descriptions and external specification files (SKILL.md). Boundary markers: None explicitly defined for the input stage; however, the skill mandates validation of the generated output artifacts (SKILL.md). Capability inventory: The skill has the capability to write files to the local file system, execute shell commands through subprocesses, and make network requests to the Airflow API. Sanitization: The skill requires defining Pydantic models to validate the payload configuration of the generated DAGs.
Audit Metadata