openspec-ff-change
Pass
Audited by Gen Agent Trust Hub on Feb 22, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION] (SAFE): The skill utilizes the
openspecCLI to manage project state. It executesopenspec new,openspec status, andopenspec instructions. Commands are parameterized with quotes, and user-provided input is transformed into kebab-case, which mitigates shell injection risks. - [INDIRECT_PROMPT_INJECTION] (LOW): The skill processes structured data from the CLI that influences its output.
- Ingestion points: Reads JSON fields (context, rules, template, instruction) from the
openspec instructionscommand. - Boundary markers: Explicit instructions are provided to the agent to treat
contextandrulesas internal constraints rather than content to be printed. - Capability inventory: The skill can execute local
openspeccommands and perform file write operations within theopenspec/changes/directory. - Sanitization: The skill relies on structured JSON parsing and agent-level adherence to output rules to prevent malicious instructions in the artifact definitions from escalating.
Audit Metadata