Selecting Agents

Agent selection depends on task type: use technical-writer for code documentation, ultrathink-debugger for complex debugging, rust-agent for Rust development, code-review-agent for code review, and plan-review-agent for workflow review. Match the agent to your task context, not keywords, to enforce the appropriate workflow.

Technical-writer handles code-sync documentation when code changes affect docs. Retrospective-writer captures lessons learned and completed work insights. Use technical-writer for ongoing documentation needs and retrospective-writer after finishing a task or debugging session.

Yes. The `/cipherpowers:execute` command uses this Skill's decision logic to automatically dispatch the correct specialized agent based on task analysis, eliminating manual selection and streamlining complex workflows.

Keyword matching often leads to incorrect agent dispatch because task requirements involve complexity, technology, and explicit instructions beyond surface keywords. This Skill analyzes actual task context and flags red-signal scenarios to avoid naive selection errors.

Agent selection logic applies to implementation, debugging, review, and documentation tasks across all domains. The decision framework categorizes agents by function—Documentation, Debugging, Development—so the same selection principle scales whether you're working with Rust, general code, or documentation workflows.

Incorrect agent dispatch breaks workflow enforcement and wastes AI resources. This Skill prevents mis-dispatch by clarifying overlapping scenarios and identifying when task attributes—not agent mentions—determine the right choice, ensuring accurate task assignment on first dispatch.