Task-Driven Development

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and implement work from external issue bodies and comments (GitHub/GitLab). This data is untrusted and could contain malicious instructions meant to override agent behavior during the implementation phase.
  • Ingestion points: The agent is instructed to read issue bodies and comments via host CLIs (gh or glab).
  • Capability inventory: The skill allows for repository modifications (file edits), branch creation, and PR/MR management.
  • Boundary markers: There are no explicit instructions for the agent to treat issue content as data rather than instructions or to use delimiters.
  • Sanitization: No sanitization or validation of the ingested issue text is performed before the agent processes it.
  • [COMMAND_EXECUTION]: The skill utilizes standard command-line tools (git, gh, glab) to perform its primary function. These are used for intended engineering purposes such as remote host resolution, issue management, and worktree operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 10:52 AM
Security Audit — agent-trust-hub — Task-Driven Development