factory-operator
Pass
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed with a strong focus on manual oversight and durable state tracking, which significantly reduces the risk of autonomous or malicious state transitions.
- [COMMAND_EXECUTION]: The skill orchestrates workflows using the 'harness factory' CLI. These commands are integral to the skill's function and are constrained by explicit instructions to wait for process exit and avoid concurrent command execution.
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests untrusted data from external trackers (Linear/GitHub). Evidence: 1. Ingestion points: issue descriptions and comments are materialized via 'factory linear fetch' (references/linear-and-config.md). 2. Boundary markers: No explicit prompt delimiters or markers are defined in the instructions. 3. Capability inventory: The skill can perform git operations and update external trackers through the 'harness factory' CLI. 4. Sanitization: No explicit sanitization or filtering is described. The risk is mitigated by the instruction to reconcile all external data against the durable, internal event log.
Audit Metadata