openspec-new-change
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill interacts with the local environment by executing commands through the 'openspec' CLI. User-provided change names and artifact IDs are used as arguments in these shell subprocesses.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied descriptions to generate identifiers used in shell commands, creating an injection surface. 1. Ingestion points: User input for change names or build descriptions in SKILL.md. 2. Boundary markers: The skill uses double quotes to wrap shell arguments in its command templates. 3. Capability inventory: Execution of 'openspec new change', 'openspec status', and 'openspec instructions' shell commands. 4. Sanitization: The skill includes an explicit guardrail requiring the agent to validate that change names follow a kebab-case format, which serves to prevent command injection from malicious user input.
Audit Metadata