phase-orchestrator

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external implementation plans, repository instructions, and phase evidence as the primary source of truth for its actions, creating a surface for potential injection. 1. Ingestion points: The skill reads target repository instructions, multi-phase plans, and external specifications as described in SKILL.md and references/orchestration-protocol.md. 2. Boundary markers: The orchestration protocol lacks explicit structural delimiters or instructions to ignore potential commands embedded within the ingested repository data. 3. Capability inventory: The skill allows for repository modification and the execution of validation commands as part of the phase loop in references/orchestration-protocol.md. 4. Sanitization: There is no evidence of sanitization, validation, or escaping of the repository-sourced data before it influences agent behavior.
  • [COMMAND_EXECUTION]: The skill facilitates the execution of commands provided by the target repository environment or documentation. Evidence: The orchestration protocol instructs the agent to execute focused validation commands and aggregate gates, specifically mentioning the use of the hra-host-run wrapper for child commands sourced from repository instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 01:49 PM