issue-to-plan
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes local shell commands, including
git,gh,grep,awk, andsed, to manage repository state, interact with GitHub's CLI, and perform file content validation. These commands are used within the scope of a standard development workflow for issue management. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from GitHub issues via the
gh issue viewcommand. This creates a surface where a maliciously crafted issue could include instructions aimed at hijacking the agent's behavior during the planning phase. - Ingestion points: The skill reads the body and title of external GitHub issues using
gh issue view <N>(defined inSKILL.md). - Boundary markers: There are no explicit instructions or delimiters provided to the agent to treat the issue content as untrusted data or to ignore embedded instructions within the issue body.
- Capability inventory: The skill possesses capabilities to write files locally (
.superpowers/sdd/plan.md), push code to remote repositories (git push), and create pull requests and comments on GitHub (gh pr create,gh pr comment). - Sanitization: No sanitization, filtering, or escaping logic is applied to the ingested issue content before it is passed to the plan generation sub-skill.
Audit Metadata