2e81516f1f
A fourth failure of the same guard, found by review before it reached CI: the config-schema property suite builds fixtures through a getWritableTmp() helper that returns the first writable of /private/tmp, /tmp, os.tmpdir(). On macOS that is /private/tmp while os.tmpdir() is /var/folders/.../T, so five cleanup() calls in that file were refused outright. The three earlier breakages were all spellings of one root. This one is not: the suite legitimately uses more than one temp root, so the premise was wrong rather than the encoding. tmpRootCandidates() now probes the conventional system temp dirs alongside os.tmpdir(), each included only if it exists on the host, so the accepted set stays a bounded explicit list instead of growing a patch per platform. Two corrections that follow from the same review: A root that is itself a filesystem root already ends in a separator, and appending another built `//`, which only the literal `/` satisfies — TMPDIR=/ would have refused every descendant. The separator is only appended when it is not already there. The refusal messages named os.tmpdir(), which stopped being the boundary. They now name the roots actually compared against. Every failure of this guard so far was diagnosed from that message in a CI log, so it should show what was checked rather than a stale approximation of it. The symlink-escape check keeps its refusal but drops its claim. fs.rmSync does not follow a top-level symlink — it unlinks the link and leaves the target alone — so that check was never closing a live escape, and the test asserting the victim survived would have passed with the guard removed. Both now say what is true: a defense-in-depth boundary against a future change to the deletion mechanism, with only the refusal itself load-bearing. The QA path helpers were evaluated for reuse rather than keeping a third copy of path containment. They resolve against one project directory and have no multi-root or Windows short-name handling, so they are not a drop-in; noted rather than forced. Refs #3057 Co-Authored-By: Claude Opus 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.