test-discipline

Pass

Audited by Gen Agent Trust Hub on Nov 3, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze pull request diffs, which are untrusted external inputs. This introduces a potential surface for indirect prompt injection if a malicious user crafts a pull request with instructions hidden in code comments or diffs. However, the skill's instructions are purely analytical and do not leverage sensitive capabilities that could be abused for exfiltration or execution.
  • Ingestion points: Processes pull request diffs matching test-related paths and file extensions (e.g., test/**, **/*.test.ts, **/__tests__/**) as defined in the applies_to metadata.
  • Boundary markers: No explicit boundary markers or instructions to ignore embedded directives in the code are specified.
  • Capability inventory: The skill is a set of text-based review instructions; it does not request or utilize tools for network access, file system modifications, or arbitrary command execution.
  • Sanitization: No sanitization or escaping of the ingested diff content is specified in the instructions.
  • [SAFE]: The skill's metadata and origin point to a known vendor repository (github.com/thrillmade/agent-skills), which is consistent with its stated purpose. There are no signs of obfuscation, hardcoded credentials, or unauthorized persistence mechanisms.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 3, 2026, 08:26 PM