Skills
Navigate
skills/pluginagentmarketplace/custom-plugin-data-engineer/api-development/Gen Agent Trust Hub

api-development

Pass

Audited by Gen Agent Trust Hub on Sep 26, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill provides patterns for building APIs that ingest and process external data, which is a potential surface for indirect prompt injection if the agent uses these patterns to handle adversarial input without proper delimiters.
  • Ingestion points: The implement_api_development function in references/GUIDE.md accepts input_data for processing.
  • Boundary markers: None identified in the provided implementation templates.
  • Capability inventory: The skill demonstrates interaction with databases, task queues (Celery), and background tasks.
  • Sanitization: The skill explicitly promotes input validation using Pydantic in SKILL.md and PATTERNS.md, and uses yaml.safe_load in scripts/validate.py and PATTERNS.md to prevent unsafe deserialization.
  • [EXTERNAL_DOWNLOADS]: The skill provides references to official documentation for FastAPI, Pydantic, and REST API design guidelines.
  • Evidence: References to fastapi.tiangolo.com, restfulapi.net, and docs.pydantic.dev in SKILL.md.
  • [COMMAND_EXECUTION]: The skill includes documentation on running local servers and task workers for development purposes.
  • Evidence: Execution of uvicorn.run and celery_app.task demonstrated in SKILL.md code snippets.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 26, 2026, 03:49 PM