do-create-tasks

Pass

Audited by Gen Agent Trust Hub on May 9, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes user-provided PRD and Technical Specification files, which creates a surface for indirect prompt injection.
  • Ingestion points: The skill reads content from ./prds/prd-[feature-slug]/prd.md and ./prds/prd-[feature-slug]/techspec.md during Step 2.
  • Boundary markers: The instructions do not define explicit boundary markers or instruction-isolation delimiters when interpolating ingested documentation into the agent context.
  • Capability inventory: The agent has file-writing capabilities to the project directory and access to specialized tools like TaskUpdate (Claude Code) and Context7 MCP (documentation lookup).
  • Sanitization: No explicit sanitization or content validation is performed on the ingested documentation before it is used to generate tasks.
  • [COMMAND_EXECUTION]: The skill performs filesystem inspection to detect and adapt to the specific AI agent's runtime environment.
  • Evidence: Step 0 involves checking for the existence of configuration directories such as .claude/, .github/, .cursor/, and .opencode/ to determine available toolsets and capability flags.
Audit Metadata
Risk Level
SAFE
Analyzed
May 9, 2026, 10:01 PM
Security Audit — agent-trust-hub — do-create-tasks