matiaspautasso/specdriven-ai-kit
The missing workflow layer for AI-assisted software development. SDD + BDD + TDD enforced via agents, commands, and templates. Works with Claude Code, Cursor, Windsurf, and Codex.
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code
$ npx degit matiaspautasso/specdriven-ai-kit/.claude ./rig-specdriven-ai-kit # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ curl -fsSL --create-dirs -o .claude/agents/bdd-scenarios-reviewer.md https://raw.githubusercontent.com/matiaspautasso/specdriven-ai-kit/master/.claude/agents/bdd-scenarios-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/gap-reviewer.md https://raw.githubusercontent.com/matiaspautasso/specdriven-ai-kit/master/.claude/agents/gap-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/performance-reviewer.md https://raw.githubusercontent.com/matiaspautasso/specdriven-ai-kit/master/.claude/agents/performance-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/regression-reviewer.md https://raw.githubusercontent.com/matiaspautasso/specdriven-ai-kit/master/.claude/agents/regression-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/spec-analyst.md https://raw.githubusercontent.com/matiaspautasso/specdriven-ai-kit/master/.claude/agents/spec-analyst.md $ curl -fsSL --create-dirs -o .claude/agents/tdd-implementer.md https://raw.githubusercontent.com/matiaspautasso/specdriven-ai-kit/master/.claude/agents/tdd-implementer.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 7,204 rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(rm -rf *)",
"Read(**/*.pem)",
"Bash(rm -rf /)",
"Bash(git push --force:*)",
"Bash(sudo *)",
"Read(.env)",
"Bash(rm -rf /*)",
"Read(./.env.*)",
"Read(~/.aws/**)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(sudo:*)",
"Bash(git reset --hard*)",
"Bash(git reset --hard:*)",
"Read(.env.*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(rm:*)",
"Bash(git rebase *)",
"Bash(wget *)",
"Bash(npm publish:*)",
"Bash(git commit *)",
"Bash(gh pr merge *)"
]
}
} Subagents (6)
| bdd-scenarios-reviewer | Use this agent after a spec.md is approved to generate Gherkin BDD scenarios. Converts business rules and acceptance criteria into concrete, readable scenarios covering happy paths, alternative flows, |
| gap-reviewer | Use this agent before implementation begins to detect ambiguities, functional gaps, missing edge cases, business contradictions, and cross-module impact not captured in the spec. Invoke after spec.md, |
| performance-reviewer | Use this agent for features that touch database queries, data processing, API payloads, list rendering, or real-time operations. Detects potential performance regressions before they reach production. |
| regression-reviewer | Use this agent after implementation is complete to verify that the feature matches its spec, all scenarios are covered, no existing modules were unintentionally broken, and all tests pass. Invoke befo |
| spec-analyst | Use this agent to analyze a feature request and produce a complete SDD specification. Invoke when starting any new feature, bug fix with scope implications, or architectural change. Returns spec.md wi |
| tdd-implementer | Use this agent to implement a specific task following strict TDD. Invoke after tasks.md exists and a specific task is selected for implementation. The agent writes the failing test first, confirms the |
Slash commands (7)
/crear-bdd/crear-plan/crear-spec/crear-tasks/implementar-tdd/revisar-gaps/revisar-implementacion
Similar rigs
AsobaCloud/claude-code-sdlc
Claude Code SDLC enforcement template — hooks, scripts, and settings for disciplined AI-assisted development
YOLO Cowboy 3.1k tok ·
anthropics/claude-code-action
—
Orchestrator 2.3k tok ·
the-blue-coder/agentic-starter
Spec-driven development starter for AI coding agents (Claude Code, OpenCode). Commands, subagents, hooks and conventions. Bring your own stack.
Orchestrator 432 tok ·
kicchann/claude-code-config
—
Automator 10.3k tok ·