ds-intake-audit
Pass
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes a "bash_exec" tool for all shell-based operations, including git maintenance and script execution. The instructions explicitly centralize all CLI activities through this specific tool rather than native shell commands.
- [DATA_EXFILTRATION]: The skill performs extensive read operations on local workspace files, including "brief.md", "plan.md", and various directories like "artifacts/" and "experiments/". However, no patterns for external data transmission or unauthorized network access were identified.
- [PROMPT_INJECTION]: The skill contains a vulnerability surface for indirect prompt injection due to its core function of auditing untrusted project data such as reviewer comments and existing drafts.
- Ingestion points: Processes content from "brief.md", "plan.md", "status.md", "SUMMARY.md", and files within the "artifacts/", "baselines/", "experiments/", "paper/", and "reviews/" directories.
- Boundary markers: Absent. The instructions do not specify the use of delimiters or warnings to ignore instructions embedded within the audited files.
- Capability inventory: Employs high-capability tools including "bash_exec" for shell operations, "artifact." for git and repository state management, and "memory." for accessing and writing to the agent's long-term memory.
- Sanitization: No explicit sanitization, validation, or escaping of the audited content is described before it is processed by the agent.
Audit Metadata