pipeline-management
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-supplied data such as sales stage goals, exit criteria, and CRM configurations to produce GTM playbooks.
- Ingestion points: Sales process parameters and CRM field mappings provided via agent interaction and defined in SKILL.md.
- Boundary markers: No specific delimiters or instructions are provided to the agent to treat external data as potentially untrusted during processing.
- Capability inventory: The skill possesses file system read access through the provided validation script (scripts/check-output.py) and file-writing capabilities to generate artifacts.
- Sanitization: There is no evidence of content sanitization or validation of the semantics of the user-provided data before it is incorporated into the generated deliverables.
- [COMMAND_EXECUTION]: The skill includes a Python script scripts/check-output.py intended for local execution to validate the structure of generated sales deliverables.
- Evidence: The script accepts a file path as a command-line argument (sys.argv[1]) and reads the file content (path.read_text()) to verify the presence of required markdown sections. While the script only outputs the presence or absence of specific headers, it involves direct interaction with the local filesystem.
Audit Metadata