Agent-Skills.md

Agent Skills: Dev Experts

Apply opinionated developer personas for architecture decisions, production debugging, language-specific code review, comprehensive reviewer passes, and test strategy. Use when you need an architect plan, devops investigation, Rust/Python/C++ review, grumpy reviewer audit, or tester-driven test plan. Triggers: architect, devops, rust-dev, python-dev, cpp-dev, reviewer, tester, pre-merge review, refactor for maintainability.

UncategorizedID: DeevsDeevs/agent-system/dev-experts

Repository

DeevsDeevs/agent-system
DeevsDeevsLicense: Apache-2.0
205

Install this agent skill to your local

pnpm dlx add-skill https://github.com/DeevsDeevs/agent-system/tree/HEAD/dev-experts

Skill Files

Browse the full folder contents for dev-experts.

Download Skill

Loading file tree…

dev-experts/SKILL.md

Skill Metadata

Name
dev-experts
Description
"Apply opinionated developer personas for architecture decisions, production debugging, language-specific code review, comprehensive reviewer passes, and test strategy. Use when you need an architect plan, devops investigation, Rust/Python/C++ review, grumpy reviewer audit, or tester-driven test plan. Triggers: architect, devops, rust-dev, python-dev, cpp-dev, reviewer, tester, pre-merge review, refactor for maintainability."

Dev Experts

Use a single skill with persona selection. If the user specifies a persona name, use that persona. If not specified, ask which persona to use and suggest one based on the task.

Persona Map

  • architect → agents/architect.md
  • devops → agents/devops.md
  • rust-dev → agents/rust-dev.md
  • python-dev → agents/python-dev.md
  • cpp-dev → agents/cpp-dev.md
  • reviewer → agents/reviewer.md
  • tester → agents/tester.md

Workflow

  1. Read README.md for the global flow and constraints.
  2. Read the selected persona file and follow its workflow verbatim.
  3. Spawn a child agent with agent_type equal to the selected persona (architect, devops, rust-dev, python-dev, cpp-dev, reviewer, tester).
  4. Pass clear task context in the spawn message, then use wait for completion.
  5. If needed, send targeted follow-ups with send_input, then wait again.
  6. Synthesize the child agent output for the user and include concrete file references.
  7. If the user asks for "refactor for maintainability", follow the persona's refactoring mode and create the plan file as described.
  8. If the task is ambiguous across personas, ask a single clarifying question and proceed.