procore-core-workflow-b

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill interacts with external data provided by the Procore API, which represents a surface for indirect prompt injection if the project data contains malicious instructions.\n
  • Ingestion points: Data retrieved from requests.get in SKILL.md is processed and printed.\n
  • Boundary markers: None present in the code snippets to delimit API-sourced text.\n
  • Capability inventory: The skill is configured with Read, Write, Edit, and Bash tools in the frontmatter.\n
  • Sanitization: No explicit sanitization or filtering of the API-returned content is performed in the provided workflow snippets.\n- [SAFE]: All API calls target the official Procore developer platform, a well-known construction technology service. The code uses the standard Python requests library to perform routine project management tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 09:56 AM
Security Audit — agent-trust-hub — procore-core-workflow-b