Agent skill · Code Review & Quality

build

Workflow orchestrator that chains existing skills for feature development

parcadeigithub.com/parcadeiGitHub ↗
claude-codeMIT
Install
npx skills add parcadei/Continuous-Claude-v3 --skill build --agent claude-code

Same command for any agent — swap --agent for codex, cursor, copilot.

Facts
Files in the skill folder: 1
SKILL.md size: 21 KB
Bundled scripts: none
Path: .claude/skills/build/SKILL.md
Open the folder on GitHub →
Where it comes from
Stars: 3,879
Language: Python
Read our review of the source →

Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.

Review
written from the skill's own SKILL.md · Aug 5, 2026

What it does

You are a workflow orchestrator that chains existing skills for feature development. You coordinate the execution of multiple skills in sequence, passing handoffs between them and pausing for human checkpoints at phase boundaries.

How it works

  • Invocation through the command "/build <mode> [options] [description]".
  • If the user provides no full arguments, you guide them through a multi-phase question flow to determine configuration:
    • Phase 0: Workflow selection with several modes (Greenfield, Brownfield, TDD, Refactor) or a help option.
    • Phase 1: Capture project context (New feature, Adding to codebase, Refactoring).
    • Phase 2: Clarify requirements (clear spec, rough idea, exploring possibilities).
    • Phase 3: Development approach (Just implement, Write tests first, Validate plan).
    • Phase 4: Post-implementation (auto-commit, PR description, leave files changed).
    • Phase 5: Feature description input.
  • The system maps selections to a Mode (greenfield, brownfield, tdd, refactor) and builds a chain of skills to run, such as discovery-interview, plan-agent, validate-agent, implement_plan, commit, describe_pr, etc., applying options like --skip-discovery, --skip-validate, --skip-commit, --skip-pr to modify the chain.
  • Step-by-step orchestration is described, including handoff directories and how to create orchestration state files, with checkpoints after key phases.
  • The orchestration includes concrete Task prompts for each subagent, with specific prompts and outputs such as specs, plans, validations, and implementation handoffs. It defines a detailed Step 1: Setup and Step 2: Execute Chain, plus detailed Phase Execution Pattern and Task structures.

When to use it

Use when you want to orchestrate feature development by chaining predefined skills in a controlled flow, with explicit human checkpoints and configurable mode (greenfield, brownfield, tdd, refactor). It is invoked via "/build" and supports interactive phase-by-phase configuration.

What it can touch

  • Handoff and work artifacts are created in: thoughts/shared/handoffs/<session>/ and related subpaths such as thoughts/shared/specs/, thoughts/shared/plans/, thoughts/shared/research/.

Caveats

  • License indicated: MIT.
  • It relies on a multi-phase interactive process and expects user input to proceed through phases.
  • The workflow can be adjusted by options to skip discovery, validation, commit, or PR description, potentially affecting artifact creation and validation steps.
From the SKILL.md

# Build - Workflow Orchestrator You are a workflow orchestrator that chains existing skills for feature development. You coordinate the execution of multiple skills in sequence, passing handoffs between them and pausing for human checkpoints at phase boundaries. ## Invocation ``` /build <mode> [options] [description] ``` ## Question Flow (No Arguments) If the user types just `/build` with no or partial arguments, guide them through this question flow to infer the right configuration. Use AskUserQuestion for each phase. ### Phase 0: Workflow Selection ```yaml question: "What would you like to do?" header: "Workflow" options: - label: "Help me choose (Recommended)" description: "I'll ask a few questions to pick the right workflow" - label: "Greenfield - new feature" description: "Chain: discovery → plan → validate → implement → commit → PR" - label: "Brownfield - existing code" description: "Chain: onboard → research → plan → validate → implement" - label: "TDD - test-first" description: "Chain: plan → test-driven-development → implement" - label: "Refactor - improve structure" description: "Chain: impact analysis → plan → TDD → implement" ``` **Mapping:** - "Help me choose" → Contin

What's inside
Steps it walks through
  1. Invocation
  2. Question Flow (No Arguments)
  3. Phase 0: Workflow Selection
  4. Phase 1: Project Context
  5. Phase 2: Requirements Clarity
  6. Phase 3: Development Approach
  7. Phase 4: Post-Implementation
  8. Phase 5: Description
  9. Summary Before Execution
  10. Modes
  11. Options
  12. Handoff Directory
  13. Orchestration Process
  14. Step 0: Parse Arguments
Commands it runs
mkdir -p "thoughts/shared/handoffs/$SESSION"
cat > "thoughts/shared/handoffs/$SESSION/orchestration.yaml" << EOF
Run impact analysis on the function/module being refactored
tldr impact <target> src/ --depth 3 > thoughts/shared/handoffs/<session>/impact-analysis.json
Also run architecture analysis
tldr arch src/ > thoughts/shared/handoffs/<session>/architecture.json
More from Continuous-Claude-v3
All skills →
About this skill
What does the build skill do?

Workflow orchestrator that chains existing skills for feature development

How do I install it?

Run `npx skills add parcadei/Continuous-Claude-v3 --skill build --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.

Where does this skill come from?

From parcadei/Continuous-Claude-v3, a repository with 3,879 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.

Is a popular skill a good skill?

Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.

Keep going