ds-context-reset
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill describes a legitimate development workflow for state persistence across sessions.
- [INDIRECT_PROMPT_INJECTION]: The workflow involves ingesting data from a local file to guide agent behavior. Ingestion points: Reads PROGRESS.md and git log -5 (SKILL.md). Boundary markers: None specified. Capability inventory: The agent executes tasks and acceptance checks based on the file content. Sanitization: No explicit sanitization mentioned as the agent is recording its own verified steps.
- [COMMAND_EXECUTION]: The agent is instructed to run commands for verification and task progression as recorded in the local progress file, which is a standard part of the intended development workflow.
Audit Metadata