What problem does it solve?
Manually selecting and coordinating the right AI agents for complex, multi-part tasks is time-consuming and prone to error, especially when work spans multiple projects or requires specialized capabilities. This skill eliminates that overhead by automatically analyzing your work context and building an optimal agent dispatch plan.
Core Features & Use Cases
- Intelligent Agent Mapping: Automatically matches task types (adversarial review, persona simulation, code validation, etc.) to the most capable available agents from your FH skill pool and installed plugins.
- Capability Fit Validation: Scores each agent against subtask requirements to avoid mismatches, such as assigning read-only audit agents to tasks that require file edits, and triggers plugin recommendations for uncovered capability gaps.
- Built-in Guardrails: Includes clarification protocols for ambiguous requests, destructive action warnings, tiered quality gates (M/S/R classification), and Wave-based execution with fan-in result integration to ensure safe, high-quality outcomes.
- Use Case Example: For a cross-project code review task requiring adversarial validation, persona testing, and fact-checking, the skill automatically composes the right agents, runs independent tasks in parallel, and integrates results into a prioritized action report.
Quick Start
Use the agent-composer skill to compose and run the optimal set of agents for my cross-project code review task that requires adversarial validation and persona simulation.
