ns-postgres-rag

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions are strictly limited to architectural analysis and report generation. It explicitly forbids the implementation of application code, delegating those tasks to a specialized harness, which maintains a clear separation of concerns.
  • [INDIRECT_PROMPT_INJECTION]: The skill includes an inventory phase (Gate 1) that involves reading untrusted content from the user's repository, such as DDL, SQL migrations, and data samples.
  • Ingestion points: Gate 1 Corpus Inventory scans repository files (DDL, SQL, migrations) and data samples to assess volume and structure.
  • Boundary markers: The skill does not define explicit delimiters for external content but operates within a strictly defined report template.
  • Capability inventory: The skill is restricted to generating a Markdown-based 'Retrieval Design Report'. It does not contain capabilities for network requests, shell command execution, or dynamic code evaluation.
  • Sanitization: No explicit sanitization of repository content is mentioned, but the risk is minimal as the agent's output is limited to architectural documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 10:21 PM
Security Audit — agent-trust-hub — ns-postgres-rag