openspec-continue-change
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to interact extensively with the
openspeccommand-line interface to manage workflows. - Evidence: Executes
openspec list,openspec status, andopenspec instructionsvia the shell to retrieve JSON data and update progress. - [INDIRECT_PROMPT_INJECTION]: The skill ingests instructions, templates, and rules from the output of an external command (
openspec instructions), which it then uses to generate file content. - Ingestion points: JSON data from
openspec instructions <artifact-id> --change "<name>" --json, specifically theinstruction,template,context, andrulesfields inSKILL.md. - Boundary markers: The skill includes explicit instructions for the agent to use
contextandrulesas constraints but not to copy them into the output files, acting as a partial boundary for data processing. - Capability inventory: The skill has the ability to read arbitrary dependency files, write to paths specified by the CLI (
resolvedOutputPath), and execute shell commands. - Sanitization: There is no evidence of sanitization or validation performed on the strings returned by the CLI before they are interpolated into the agent's generation task.
Audit Metadata