work-card

Pass

Audited by Gen Agent Trust Hub on Oct 30, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from external sources which could be manipulated to inject malicious instructions into the agent's context.
  • Ingestion points: The skill reads task definitions from docs/task-breakdown/sprint-N.md and reads GitHub issues using gh issue view <number> (SKILL.md, Step 1).
  • Boundary markers: No explicit markers, delimiters, or system instructions are provided to the agent to treat the content retrieved from GitHub or local markdown files as untrusted data meant for processing rather than instruction following.
  • Capability inventory: The skill possesses significant capabilities, including shell command execution (git, gh, python3), file system modification, and the ability to trigger other specialized skills like tdd and security-audit (SKILL.md, Steps 3-6).
  • Sanitization: The instructions do not mandate any validation or sanitization of the card description, AC (Acceptance Criteria), or issue content before the agent begins the build and check phases.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 30, 2026, 06:40 AM