菜单
Pull later framework improvements (process, gates, skills, hooks, scripts) from the upstream template into this project — without touching project content. Use when the template repo has shipped updates you want to adopt; it syncs only framework-owned paths and stages them for review.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Record a new architectural decision, or supersede an existing one, as a plain-markdown ADR under docs/adr/. Use when making a significant or hard-to-reverse design decision, or when an accepted decision must change (which requires a NEW superseding ADR — accepted ADRs are immutable). Handles the frontmatter schema, numbering, and the additive supersede workflow (accepted ADRs are never edited — reverse links are derived).
Set up and operate the git-remote sync of the agent's machine-local memory directory. Use to configure the memory remote the first time, to push or pull memory manually, to onboard a new machine, or to resolve a memory merge conflict. The memory_push Stop hook handles routine per-turn pushes automatically once setup is done.
Author or revise a work item's spec at docs/spec/<id>.md before any code is written, then run it through the spec gate. Use when starting a new work item, when a spec needs revision after a spec-gate rejection, or when acceptance criteria must be defined. Produces a fully-signed spec — Goal, Non-goals, evidence-cited claims, stable spec-item IDs, and frozen executable acceptance criteria.
Resolve an uncertainty during spec authoring by running a throwaway investigation and recording a concluded docs/spike/SPIKE-<n>.md evidence artifact. Use when a spec or ADR makes a design claim that is not yet backed by a concluded spike, a cited code path, or a doc — the spec gate rejects uncited claims and open questions.
| name | update-from-template |
| description | Pull later framework improvements (process, gates, skills, hooks, scripts) from the upstream template into this project — without touching project content. Use when the template repo has shipped updates you want to adopt; it syncs only framework-owned paths and stages them for review. |
A GitHub template is a one-time copy with no upstream link, so framework improvements don't arrive on their own. This skill performs a deliberate, reviewable sync: it checks out the framework-owned paths from the template remote over your copy and leaves everything project-owned untouched.
.claude/, .codex/, .agent/, .github/,
docs/process.md, docs/coding-standards.md, bootstrap.*, .gitattributes,
AGENTS.md, CLAUDE.md.pixi.toml, docs/{spec,adr,spike}/*, your source,
README.md, and docs/project-notes.md — which is exactly why project-specific
agent instructions live in docs/project-notes.md, so AGENTS.md / CLAUDE.md stay
syncable.Commit or stash framework changes first. The helper refuses if any framework file is dirty (the checkout would collide). Uncommitted project work is fine.
Run the sync — it stages the updated framework files; it does not commit:
python .agent/scripts/update_from_template.py <template-git-url> # first time — adds a `template` remote
python .agent/scripts/update_from_template.py # thereafter — reuses it
(--ref <branch> to track a non-main branch.)
Review the staged diff — this is framework code that will run in your repo.
git diff --cached. The sync only adds / updates; it does not delete files
the upstream removed, so scan git status for now-stale framework files and delete
them by hand if the template dropped them.
Regenerate, re-verify, then commit. The .claude/ + .codex/ agent surface is
gitignored build output, so rebuild it from the synced .agent/ with
python .agent/scripts/gen_agents.py. Then run pixi run check (and re-run the gates
if you're mid-work), and commit the result as a framework-only change.
The same script syncs .codex/ (it's a framework path). Trust the repo so Codex
loads the refreshed .codex/ layer. See .codex/README.md.