Claude Code Sdlc

Claude Code SDLC

Turn Claude Code into a full software development team.

16 specialized AI agents. Documentation-first. TDD. Quality gates. Hardened against Claude Code's known limitations.

Claude Code SDLC is an open-source (MIT) plugin that turns Anthropic's Claude Code into a
16-agent software development team enforcing documentation-first planning, test-driven
development, and 9 pre-merge quality gates. It installs with a single curl command and hardens
Claude Code against its documented failure modes: context loss, silent read truncation, and
unverified success reports.

Quick start

Then confirm it landed:

Open a new session in any project and the harness is active — user-scope enablement loads the
plugin everywhere (measured on Claude Code 2.1.237: all 16 agents resolve and hooks fire in a
fresh directory with no per-project setup).

Why

Claude Code out of the box:

- Writes code before understanding requirements
- Skips tests or writes them as an afterthought
- Reports success when the code doesn't compile — "bytes hit disk" is its success metric
- Silently loses context at ~167K tokens, then edits against stale memory
- Misses references during renames — grep is text matching, not an AST
- Truncates file reads at 2,000 lines and search results at ~50K chars without warning

Anthropic's own Claude Code best-practices guide
puts the core problem plainly: "Claude stops when the work looks done. Without a check it can run,
'looks done' is the only signal available, and you become the verification loop." This harness
operationalizes that advice: every plan slice carries Files:, Changes:, Verify:, and
Done when: fields, and a verifier agent re-checks wiring goal-backward after implementation.

What This Fixes

- Documentation-first — PRD, use cases, test cases…