-
Notifications
You must be signed in to change notification settings - Fork 1.4k
Description
Display Name
ADD (Agent Driven Development)
Category
Agent Skills
Sub-Category
General
Primary Link
https://github.com/MountainUnicorn/add
Author Name
MountainUnicorn
Author Link
https://github.com/MountainUnicorn
License
MIT
Other License
No response
Description
A structured SDLC methodology plugin for Claude Code where AI agents are first-class development team members. Coordinates specialized agent teams — test-writers, implementers, reviewers, deployers — using spec-driven TDD with independent verification. Features three engagement modes (guided/balanced/autonomous), away mode for autonomous work, a maturity lifecycle dial (POC→GA), 5-level quality gates, and a 3-tier cross-project learning system. Zero dependencies — pure markdown and JSON.
Validate Claims
Install via /plugin marketplace add MountainUnicorn/add then /plugin install add@add-marketplace. Run /add:init in any project to see the structured interview scaffold a full project structure in ~5 minutes. Run /add:spec "user auth" to see it generate a complete feature specification with acceptance criteria and test cases.
Specific Task(s)
- Bootstrap a new project with structured PRD interview: /add:init
- Generate feature specifications with acceptance criteria: /add:spec
- Execute full TDD cycles with coordinated agent teams: /add:tdd-cycle
- Run 5-level quality gates: /add:verify
- Work autonomously while human is away: /add:away
Specific Prompt(s)
- /add:init — 5-minute interview that scaffolds project structure, PRD, and config
- /add:spec "feature name" — structured interview generating complete feature spec
- /add:tdd-cycle specs/feature.md — dispatches test-writer, implementer, reviewer, verifier
agents - /add:away "back in 2 hours" — autonomous work plan with defined boundaries
- /add:retro — agent-to-human retrospective that promotes learnings across projects
Additional Comments
No response
Recommendation Checklist
- I have checked that this resource hasn't already been submitted
- My resource provides genuine value to Claude Code users, and any risks are clearly stated
- All provided links are working and publicly accessible
- I am submitting only ONE resource in this issue
- I understand that low-quality or duplicate submissions may be rejected