write-plan

Pass

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides a legitimate framework for project management and task decomposition within a 1C development environment.
  • [COMMAND_EXECUTION]: The skill instructions reference the use of specialized 1C development tools and validation scripts (e.g., 1c-syntax-checker, /cf-validate, /db-update). These tools are used for static analysis and database updates as part of the intended development workflow.
  • [PROMPT_INJECTION]: The skill processes untrusted user data from design.md files to generate implementation plans, presenting a surface for indirect prompt injection. This is an inherent risk of task-planning skills and is mitigated by the skill's design, which suggests human review of the generated plan.
  • Ingestion points: openspec/changes/<feature>/design.md (read via the Read tool)
  • Boundary markers: Absent; content from the design file is processed directly without explicit delimiters.
  • Capability inventory: The skill has the capability to write task lists (Write) and execute validation commands via shell (Bash).
  • Sanitization: Absent; the skill relies on the agent to interpret the documentation content correctly.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 24, 2026, 06:14 PM
Security Audit — agent-trust-hub — write-plan