ln-200-scope-decomposer

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill acts as a pure coordinator, delegating work to trusted local sub-skills using relative paths (e.g., ../ln-210-epic-coordinator/SKILL.md). All external references and tools originate from the same vendor/author namespace.- [PROMPT_INJECTION]: The skill has a potential surface for indirect prompt injection as it processes project documentation like requirements.md and architecture.md through its sub-skills. * Ingestion points: Documentation files (requirements.md, architecture.md, tech_stack.md) read during decomposition phases. * Boundary markers: No explicit boundary markers or delimiters for untrusted content are defined in the orchestrator logic. * Capability inventory: Sub-skills can create documents and Linear issues. * Sanitization: No explicit content sanitization or validation is specified at this orchestration level. This risk is inherent to the skill's primary purpose and is managed via downstream coordinator logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 01:43 PM