brainstorm-to-issue
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes the GitHub CLI (
gh) to create repository labels and issues. It passes context-derived data, such as issue titles and body content, as command-line arguments. This behavior is standard for the skill's purpose and includes a requirement to show the draft to the user and obtain confirmation before the command is executed. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from previous brainstorming conversations to generate structured issue reports.
- Ingestion points: Brainstorming conversation history (as described in
SKILL.md). - Boundary markers: No explicit input delimiters are defined; however, the skill mandates a human review of the generated output (Step 4 in
SKILL.md) before any external actions are taken. - Capability inventory: The skill has the capability to execute shell commands (
gh issue create) and write files to the local.scratch/directory as specified inSKILL.md. - Sanitization: The skill relies on the agent to extract and summarize content rather than performing programmatic sanitization.
Audit Metadata