grace-refresh
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
graceCLI tool to performlintandstatusoperations on the project root as specified in the workflow section of SKILL.md. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from repository files and metadata which creates a potential surface for indirect instructions.
- Ingestion points: SKILL.md defines a workflow that reads
.grace/graph/index.xml,.grace/verification/index.xml, and file-local markers like module contracts and tests. - Boundary markers: Absent. There are no explicit instructions or delimiters provided to ensure the agent ignores malicious content within these files.
- Capability inventory: The skill is capable of executing shell commands (
grace lint,grace status) and reporting findings to the user. - Sanitization: Absent. No filtering or validation of the ingested file content is mentioned.
Audit Metadata