一键导入
save-progress
Save progress checkpoint — commits work in progress, updates the active plan's checkboxes, and writes a numbered session summary to sessions_dir
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。
菜单
Save progress checkpoint — commits work in progress, updates the active plan's checkboxes, and writes a numbered session summary to sessions_dir
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
| name | save-progress |
| description | Save progress checkpoint — commits work in progress, updates the active plan's checkboxes, and writes a numbered session summary to sessions_dir |
| metadata | {"phase":"checkpoint","input":"no arguments — reads session context, git state, and active plan automatically","output":"WIP commit, updated plan file, session summary saved to sessions_dir as NNN_feature.md","dependencies":"the tool's workspace file, if configured via central-workspace — informational only, not directly loaded or searched by this skill"} |
Create a comprehensive progress checkpoint when pausing work on a feature.
git status to check for uncommitted changesCommit meaningful work:
git status
git diff
git add [specific files]
git commit -m "WIP: [Feature] - [Current state]"
If changes are not commit-ready, note:
If working from a plan, append a progress checkpoint to it:
## Progress Checkpoint - [Date Time]
### Work Completed This Session
- [x] Specific task completed
- [x] Another completed item
- [ ] Partially complete task (50% done)
### Current State
- **Active File**: `path/to/file:123`
- **Current Task**: [What you were doing]
- **Blockers**: [Any issues encountered]
### Local Changes
- Modified: `file1` - Added validation logic
- Modified: `file2` - Partial refactor
- Untracked: `test.tmp` - Temporary test file
### Next Steps
1. [Immediate next action]
2. [Following task]
3. [Subsequent work]
### Context Notes
- [Important discovery or decision]
- [Gotcha to remember]
- [Dependency to check]
### Commands to Resume
```bash
# To continue exactly where we left off:
git status
# invoke implement skill with the plan file path
---
## Step 4 — Create Session Summary
Use `sessions_dir` (default: `docs/sessions/`) and `research_dir` (default: `docs/research/`) from `# WORKSPACE`.
Check existing files in `sessions_dir` to determine the next sequential number (001, 002, …).
Save as `NNN_feature.md` with this structure:
```markdown
---
date: [ISO timestamp]
feature: [Feature name]
plan: [path to plan file]
research: [path to research file if exists]
status: in_progress
last_commit: [git hash]
---
# Session Summary: [Feature Name]
## Session Duration
- Started: [timestamp]
- Ended: [timestamp]
- Duration: [X hours Y minutes]
## Objectives
- [What we set out to do]
## Accomplishments
- [What was actually completed]
## Discoveries
- [Important findings]
## Decisions Made
- [Architecture choices]
- [Trade-offs accepted]
## Open Questions
- [Unresolved issues]
## File Changes
```bash
git diff --stat HEAD~N..HEAD
To continue this work:
[plan path][research path if exists]
---
## Step 5 — Clean Up
1. Commit all meaningful changes using the `commit` skill
2. Update the todo list to reflect the saved state
3. Present summary to the user:
✅ Progress saved successfully!
📁 Session summary: [sessions_dir][NNN_feature.md] 📋 Plan updated: [plan path] 💾 Commits created: [list]
To resume: invoke the resume-work skill with [sessions_dir][NNN_feature.md]
---
## Guidelines
- **Always commit meaningful work** — don't leave important changes uncommitted
- **Be specific in notes** — future sessions need clear context, not summaries
- **Include commands** — make resuming as close to copy-paste as possible
- **Document blockers** — explain why work stopped, not just that it stopped
Create GitHub pull requests — picks the most token-efficient available method (gh CLI, GitHub MCP, or git push), and sources the PR body from an existing spec/plan instead of re-analyzing the diff
Prime — load project context by reading docs, the docs_context layer, core docs, and TODO before starting any task
Feature planning — research codebase, design options, write a structured spec/implementation plan to the project specs directory
Generate feature documentation by analyzing code changes and specifications — creates markdown docs in docs/core/ with conditional context
Bootstrap any project's AI workspace — detect the current tool, extract hardcoded paths from skill/command/agent files, write a central workspace.md, update all files to use config keys
Build or refresh the codebase graph indexes (codebase-memory-mcp structural index + graphify semantic graph) that other skills read from. The single writer — idempotent, incremental, safe to call repeatedly.