develop

Enforce a plan-before-build workflow with approval gating and verification.

Updated May 21, 2026
One-click install
npx skills add https://github.com/mica-agent/mica-solar-system --skill develop-mica-agent
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: develop
Source: https://github.com/mica-agent/mica-solar-system/tree/main/.qwen/skills/develop
Command: npx skills add https://github.com/mica-agent/mica-solar-system --skill develop-mica-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It prevents agents from skipping required build steps by enforcing a consistent plan-before-build workflow, including dependency research, canvas/spec correctness, an explicit approval gate, and verification/document reconciliation.

Core Features & Use Cases

  • Universal build flow gate: Orchestrates planning, dispatch, and correctness invariants for build-shaped requests (canvas cards, standalone tools, and doc-only specs).
  • Dependency and API grounding: Triggers research when non-trivial domain work is involved and loads the canvas handbook when building card classes to reduce wrong API usage.
  • Approval-gated execution: Writes the spec, then explicitly pauses for the user’s next message before moving to implementation—avoiding silent drift from the user’s intent.

Quick Start

Tell your agent to use the develop skill to plan and spec a new build artifact, then ask you for review before it writes any code.

Frequently Asked Questions about develop

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I enforce a plan-before-build workflow for agent-driven software builds?

A plan-before-build workflow enforces planning, dependency research, spec generation, and approval gating before any code dispatch occurs. It mandates an initial todo_write call to capture the build steps, preventing agents from skipping required structural phases.

Can I gate code generation behind an explicit approval step for canvas cards?

Yes, approval gating pauses execution after spec generation and before implementation. The agent writes the build spec, then explicitly waits for your next message to proceed, preventing silent drift from your intended canvas card design.

What's the best way to verify build correctness and maintain doc consistency after task decomposition?

Verify build correctness by requiring render_capture or equivalent testing verification after dispatch. You maintain doc consistency by mandating document reconciliation steps alongside the render verification to ensure outputs match the original spec.

Does this build orchestration approach work for standalone programs and doc-only artifacts?

Yes, build orchestration applies universally to canvas cards, standalone programs, and doc-only artifacts. It routes build-shaped requests through dependency research, decomposition, and verification invariants regardless of the artifact type.

Why do agents skip dependency research and API grounding during complex builds?

Agents skip dependency research when a workflow lacks explicit grounding triggers. Enforcing a universal build flow gate triggers research for non-trivial domain work and loads the card-class-handbook for canvas builds to reduce wrong API usage.

When do I need spec generation and approval gating for build-shaped requests?

You need spec generation and approval gating whenever a build-shaped request requires decomposition and dispatch. This process ensures the agent writes a spec, pauses for your review, and only then proceeds to implementation to avoid incorrect outputs.