ship-cycle
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a wide range of Git and GitHub CLI (gh) commands to automate development workflows, including issue creation, branch management, pull requests, tagging, and release publishing. These operations are essential to the skill's primary function and are performed following project conventions.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface by ingesting untrusted data from external sources (such as sweep reports or manually pasted notes) at Step 4 to generate issue descriptions, PR bodies, and handoff notes intended for future AI sessions. \n
- Ingestion points: Step 4 (Intake) in
SKILL.mdaccepts external improvement lists and reports.\n - Boundary markers: The skill uses structured templates with English headings and checkboxes (defined in
issue-body-template.md) to provide structural delimiters.\n - Capability inventory: The skill possesses the capability to commit code, create public issues, and publish releases via the
Tasktool and shell commands (Steps 8-11 inSKILL.md).\n - Sanitization: A mandatory human-in-the-loop confirmation gate at Step 7 (Confirm plan) requires the operator to review and approve the drafted issues and milestones before execution.
Audit Metadata