Tom Boucher f2c077df38 chore(#2387): refactor CONTEXT.md legacy content + add glossary drift gate (#2391)
* chore(#2387): refactor CONTEXT.md legacy content + add glossary drift gate

Apply the audit-and-enforce concept from the ADR index (#2356) to CONTEXT.md:
correct stale facts, and add a CI gate so the machine-verifiable claims can't
silently re-rot.

CONTEXT.md was entirely hand-maintained with nothing checking its claims against
the shipped tree, so it had rotted. An audit against live code (Memtrace +
filesystem + gh), each finding adversarially re-verified, drove 38 factual
corrections + 1 surfaced by the new gate:

- Dead references: Package Identity named @opengsd/get-shit-done-redux (package
  is @opengsd/gsd-core); Shell Command Projection named run-git/run-npm/run-tool
  (real exports execGit/execNpm/execTool); a partial docs/adr/1606 ref; retired
  sdk/ framing.
- Superseded facts: allRuntimes 15 -> 17 (pi #2102, zcode); "seven nested-loader
  runtimes" -> five (claude reverted flat #924, antigravity flat); stacked-PR
  examples rebasing onto main -> next; QUOTA_SENTINELS precedence corrected to
  match src/agent-command-router.cts.
- Drifted CONTRIBUTING.md line citations refreshed.

Per CONTRIBUTING.md:179, only stale FACTS were corrected -- no maintainer intent,
lesson, or opinion was rewritten, and the append-only session log is untouched
except one dated in-place superseding note. The three tests that assert on
CONTEXT.md content (phase6-capstone-conformance, tracer-bullet,
external-job-waiting) keep all their anchors.

New scripts/check-glossary-refs.cjs (--check, wired into lint:generated-sync):
- Check A: every backticked file reference under a TRACKED_PREFIXES allowlist
  resolves on disk. Generated gsd-core/bin/lib/*.cjs (77 refs, gitignored),
  ~/-paths, .planning/, and bare filenames are deliberately skipped so a clean
  CI checkout never false-fails.
- Check B: the allRuntimes count + member set in the glossary prose match
  bin/install.js's allRuntimes literal (drifts on every runtime addition).
tests/check-glossary-refs.test.cjs covers both, including the false-positive
guard that a missing bin/lib/*.cjs ref does NOT trip the gate.

Closes #2387

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix(#2387): confine glossary-gate file refs to ROOT (no `..` traversal)

Pre-PR security review finding (low): extractTrackedRefs fed tokens straight to
fs.existsSync(path.join(ROOT, token)), and PATH_TOKEN_RE admits `.` in a segment,
so a CONTEXT.md token like `src/../../../etc/passwd` passed the `src/` prefix
check and normalized to an out-of-tree absolute path — turning the doc lint into
a filesystem-existence oracle on the CI host (existsSync only; CONTEXT.md is a
trusted committed file, hence low severity, but a defense-in-depth gap).

Add isWithinRoot() confinement in extractTrackedRefs: a token is dropped unless
path.resolve(ROOT, token) stays within ROOT. A CONTEXT.md reference is always a
plain in-repo path, so a `..` escape is never legitimate. Regression test asserts
a `..`-bearing token is skipped and never named in output.

Refs #2387

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix(#2387): drop legacy `get-shit-done` name from a CONTEXT.md defect entry

CI lint-legacy-dir-name failed: the line-928 upstream-issue re-point I applied
wrote the historical provenance as "gsd-build/get-shit-done#3545", and
scripts/lint-legacy-dir-name.cjs forbids the legacy `get-shit-done` name. Reword
to "moved from #3545 in the predecessor repo" — same provenance, no legacy name.

Caught by `npm run lint:ci` (the CI lint chain), which I had not run locally —
lint:generated-sync + eslint do not include lint-legacy-dir-name.

Refs #2387

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-17 19:24:28 -04:00

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.

npm version npm downloads Tests Discord GitHub stars License


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:

  1. Discuss — capture implementation decisions before anything is planned
  2. Plan — research, decompose, and verify the plan fits a fresh context window
  3. Execute — run plans in parallel waves; each executor starts with a clean 200k-token context
  4. Verify — walk through what was built; diagnose and fix before declaring done
  5. 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

Star History Chart

License

MIT License. See LICENSE for details.


Claude Code is powerful. GSD Core makes it reliable.

Description
No description provided
Readme MIT 79 MiB
Languages
JavaScript 82.2%
TypeScript 17.4%
Shell 0.4%