delivery-engine

Warn

Audited by Gen Agent Trust Hub on Oct 7, 2026

Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The validation script tests/run-validation.py uses exec() and compile() to load and execute code from scripts/validate-output.py. The use of dynamic execution based on computed file paths, even for local files, is a risk factor.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided project specifications and briefs to generate structured task lists, which introduces a surface for indirect prompt injection.
  • Ingestion points: The agent reads and processes user-supplied scope documents or feature briefs as input for task decomposition as documented in SKILL.md.
  • Boundary markers: No explicit delimiters or instructions are provided to help the agent distinguish between its own logic and content within the input specification.
  • Capability inventory: The skill uses the read_file tool to access file content.
  • Sanitization: No sanitization or validation of the input content is implemented to detect or strip malicious prompt instructions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 7, 2026, 11:26 AM