747a3730d4
* test(#4268): harden reworded-restatement and backend-predicate-divergence detection tests/tdd-single-statement.test.cjs's restatesCycle() keyed on the exact literal `commit: `test({phase}-{plan})`` substring, so a reworded restatement of the RED/GREEN/REFACTOR procedure shipped green. Adds restatesCycleStructurally(), a structural (span + list-marker) detector that stays linear-scan (per the #4228 catastrophic-backtracking incident this must not reintroduce) and is proven, empirically, to flag a paraphrased multi-step fixture while not flagging the real compact citations in execute-plan.md and gsd-executor.md (#4267's legitimate pointers). tests/tdd-backend-wiring.test.cjs never compared the two dispatch backends' `gsd_run query phase.tdd-applicable` calls against each other, so a one-word divergence between them (e.g. a changed --pick flag in only one backend) shipped green. Adds a byte-identity assertion on the command-substitution content (normalized for the two backends' differing variable-name prefixes), proven to have teeth via a RED-first mutation check before asserting it against the real files. The third gap in #4268 (nothing proves TDD_APPLICABLE has a real definition) was already covered by this file's existing assertTddApplicableIsComputed (epic #4272 Phase 2, #4266) — verified by inspection, no new test needed. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> * test(#4268): redesign restatement detector around deferral, not length Standards+Spec review of the prior commit proved by execution that a compact, no-list-marker restatement (under the 200-char span threshold) sails past the span/list-marker-only signal. Redesigns the primary check: the actual invariant is deferral, not length — a legitimate RED/GREEN/ REFACTOR mention always names tdd.md as the authority nearby, a restatement never does. Flags when no tdd.md/canonical reference appears within a 500-char trailing window past the cycle mention, regardless of length or list-marker shape; keeps span>200 and three-distinct-list-marker-lines as secondary defense-in-depth OR-conditions. Verified independently against both real files (execute-plan.md span=10, gsd-executor.md span=14, both with a nearby deferral marker at +228/+82 chars) — no false positive, and the reviewer's exact gap class (a 189-char no-citation paraphrase) is now flagged. Also fixes: boundary coverage at the span threshold (199/200/201, isolated via a factored-out measureCycleSpan() helper), a fast-check property test proving the fix holds for arbitrary filler text, and a fragile line-match in tdd-backend-wiring.test.cjs that happened to work only because a FATAL echo message containing the same substring came later in document order than the real assignment line. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> --------- Co-authored-by: sim <sim@local> Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
GSD Core
Git. Ship. Done.
English · Português · 简体中文 · 日本語 · 한국어
A light-weight meta-prompting, context engineering, and spec-driven development system for Claude Code, OpenCode, Antigravity CLI, Kimi CLI, Kilo, Codex, Copilot, Cursor, Windsurf, and more.
What is GSD Core
GSD Core is a context-engineering and spec-driven development framework that drives AI coding agents (Claude Code, Codex, Antigravity CLI, Kimi CLI, Copilot, Cursor, and more) through a disciplined phase loop. It solves context rot — the quality degradation that accumulates as an AI fills its context window — by running all heavy research, planning, and execution work in fresh-context subagents while keeping your main session lean.
How it works
Each milestone repeats the same five-step loop, one phase at a time:
- Discuss — capture implementation decisions before anything is planned
- Plan — research, decompose, and verify the plan fits a fresh context window
- Execute — run plans in parallel waves; each executor starts with a clean 200k-token context
- Verify — walk through what was built; diagnose and fix before declaring done
- Ship — create the PR, archive the phase, repeat for the next one
Quickstart
npx @opengsd/gsd-core@latest
The installer prompts for your runtime (Claude Code, OpenCode, Antigravity CLI, Kimi CLI, Kilo, Codex, Copilot, Cursor, Windsurf, and more) and whether to install globally or locally. The installer is required for cross-runtime compatibility — do not copy files from agents/ or commands/ directly.
On another runtime or without Node.js? See Install on your runtime.
Once installed, start a new project or onboard an existing repo:
/gsd-new-project # greenfield project
/gsd-onboard # existing codebase
New here? Follow Your first project for a guided walkthrough from install to first shipped phase, or Onboarding an existing codebase for brownfield setup.
Documentation
What's new in 1.7.0 → docs/whats-new-1.7.0.md
Tutorials — learning by doing:
How-to guides — task-focused recipes:
Reference — authoritative facts:
Explanation — concepts and design decisions:
Full index: docs/README.md. Other languages: 日本語 · 한국어 · Português · 简体中文.
Why it works
Most AI-coding setups fail at scale because context bloat silently degrades output quality, there is no shared memory between sessions, and nothing verifies that code actually works. GSD Core solves all three: heavy work runs in fresh subagents, structured artifacts like STATE.md and CONTEXT.md survive session boundaries, and the verify step walks through what was built and generates fix plans before a phase is declared done. See docs/explanation/context-engineering.md for the full reasoning.
Troubleshooting? See docs/how-to/recover-and-troubleshoot.md.
Community
| Project | Platform |
|---|---|
| gsd-opencode | Original OpenCode port |
| Discord | Community support |
Star History
License
MIT License. See LICENSE for details.
Claude Code is powerful. GSD Core makes it reliable.