speckit-updater
SpecKit Safe Update
- License 许可
- MIT
- 最近更新
- 2026-11-10
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- speckit-updater
- risk
- unknown
- source
- community
- description
- SpecKit Safe Update
03 SKILL.md 正文
SpecKit Safe Update
This skill provides safe update capabilities for GitHub SpecKit installations, preserving customizations while applying template updates.
Installation: Available via plugin (/plugin marketplace add NotMyself/claude-plugins then /plugin install speckit-updater) or manual Git clone. See README.md for details.
When to Use
- You need to update or install SpecKit templates while preserving project customizations.
- You want a safe approval flow around update, rollback, or version-specific SpecKit operations.
- The task is to operate the SpecKit updater conversationally instead of running raw commands blindly.
What to do when this skill is invoked
When the user invokes /speckit-updater, you should:
-
Run the update orchestrator script without any flags (conversational mode):
pwsh -NoProfile -Command "& 'C:\Users\bobby\.claude\skills\speckit-updater\scripts\update-wrapper.ps1'" -
Parse the output for markers:
[PROMPT_FOR_APPROVAL]- Update scenario (existing SpecKit installation)[PROMPT_FOR_INSTALL]- Fresh installation scenario (no .specify/ directory)
-
For Updates (
[PROMPT_FOR_APPROVAL]marker found):- Present the Markdown summary showing:
- Current version vs. available version
- Files to update/add/remove
- Conflicts detected (if any)
- Files preserved (customized)
- Backup location
- Custom commands
- Ask the user for approval to proceed with the update
- If approved, re-run with
-Proceedflag - If declined, inform the user the update was cancelled
- Present the Markdown summary showing:
-
For Fresh Installations (
[PROMPT_FOR_INSTALL]marker found):- Present a natural installation offer to the user, such as:
- "SpecKit is not currently installed in this project. Would you like me to install it?"
- "I can install the latest SpecKit templates for you. This will create the .specify/ directory structure and download the templates from GitHub."
- Do NOT mention the
-Proceedflag to the user (this is an implementation detail) - If user approves (says "yes", "proceed", "install it", etc.), re-run with
-Proceedflag - If user declines, inform them the installation was cancelled
- Present a natural installation offer to the user, such as:
-
Execute approved action by re-running with
-Proceedflag:pwsh -NoProfile -Command "& 'C:\Users\bobby\.claude\skills\speckit-updater\scripts\update-wrapper.ps1' -Proceed"
Special cases:
- If user requests
-CheckOnly: run with that flag and show the report - If user requests
-Rollback: run with that flag and confirm restoration - If user requests specific
-Version: include that parameter
Commands
/speckit-updater
Updates SpecKit templates, commands, and scripts while preserving customizations.
Usage:
/speckit-updater- Interactive update/install with conversational approval workflow (recommended for Claude Code)/speckit-updater -Proceed- Proceed with update/install after approval (used by Claude after user confirms)/speckit-updater -CheckOnly- Check for updates without applying/speckit-updater -Version v0.0.72- Update to specific version/speckit-updater -Force- Force overwrite SpecKit files (preserves custom commands)/speckit-updater -Rollback- Restore from previous backup/speckit-updater -Auto- DEPRECATED: Use conversational workflow instead (shows warning, maps to -Proceed)
Fresh Installation (No .specify/ directory):
- First invocation shows installation offer with
[PROMPT_FOR_INSTALL]marker - Claude Code presents natural question to user (e.g., "Would you like me to install SpecKit?")
- User approves via conversational response (e.g., "yes", "proceed", "install it")
- Claude re-invokes with
-Proceedflag automatically (implementation detail hidden from user) - Script creates
.specify/structure, downloads templates, creates manifest - Exit code 0 throughout (awaiting approval is not an error)
- Consistent with update flow: both use conversational approval workflow
Process:
- Validates prerequisites (Git installed, clean Git state, write permissions)
- Loads or creates manifest (.specify/manifest.json)
- Fetches target version from GitHub Releases API
- Compares file hashes to identify customizations
- Creates timestamped backup
- Applies selective updates preserving customized files
- Opens VSCode merge editor for conflicts (Flow A: one at a time)
- Automatically invokes /speckit.constitution for constitution updates
- Updates manifest with new version
- Manages backup retention (keeps last 5)
When you invoke this command, I will:
- Execute the update-orchestrator.ps1 script
- Parse output for markers (
[PROMPT_FOR_APPROVAL]for updates,[PROMPT_FOR_INSTALL]for fresh installations) - For updates: Present Markdown summary of proposed changes
- For installations: Ask naturally if you want to install SpecKit (without mentioning
-Proceedflag) - Wait for your approval via chat conversation
- After approval: automatically re-invoke with
-Proceedflag to execute - Guide you through conflict resolution one file at a time (updates only)
- Open VSCode diff/merge tools as needed (updates only)
- Report results with detailed summary
Conversational Workflow: The skill uses a two-step approval process:
- Step 1: Outputs summary → script exits → waits for approval
- Step 2: After approval, Claude re-invokes with
-Proceed→ applies updates
Requirements:
- Git installed and in PATH
- Internet connection for fetching updates from GitHub
- Write permissions to .specify/ and .claude/ directories
- Clean or staged Git working directory
The script is located at: {skill_path}/scripts/update-wrapper.ps1 (entry point) and {skill_path}/scripts/update-orchestrator.ps1 (main logic)
Entry point command:
pwsh -NoProfile -Command "& '{skill_path}/scripts/update-wrapper.ps1' [parameters]"
Note: Both PowerShell-style (-CheckOnly) and Linux-style (--check-only) flags are supported via the wrapper script.
Features
- Customization Preservation: Automatically detects and preserves user customizations using normalized file hashing
- Intelligent Conflict Resolution: Guides through conflicts one-at-a-time with 4 options: merge editor, keep mine, use new, skip
- Version Tracking: Maintains
.specify/manifest.jsonwith file hashes, version info, and backup history - Automatic Backups: Creates timestamped backups in
.specify/backups/with automatic retention management - Fail-Fast with Rollback: Automatically rolls back on any error, restoring pre-update state
- Dry-Run Mode:
--check-onlyshows exactly what would change without applying updates - Constitution Integration: Notifies when constitution template has updates (run
/speckit.constitution) - Custom Command Safety: User-created commands never overwritten, even with
--force
Architecture
Modules
- HashUtils: Normalized hashing (handles line endings, trailing whitespace, BOM)
- VSCodeIntegration: Context detection, Quick Pick, diff/merge editor integration
- GitHubApiClient: GitHub Releases API interaction (unauthenticated, 60 req/hour)
- ManifestManager: Manifest CRUD operations with caching
- BackupManager: Backup creation, restoration, and retention management
- ConflictDetector: File state analysis and conflict detection
Workflow
- Prerequisites validation (critical checks must pass, warnings allow continuation)
- Manifest loading/creation (safe default: assume all files customized if no manifest)
- GitHub API query for target version
- File state analysis (6 actions: add/remove/merge/preserve/update/skip)
- User confirmation with change preview
- Backup creation (timestamped, excludes backups directory)
- Selective file updates (fail-fast with automatic rollback)
- Conflict resolution (Flow A: one-at-a-time, VSCode merge editor)
- Manifest update (version, file hashes, customization flags)
- Backup cleanup (keep 5 most recent, requires confirmation)
- Detailed summary display
Exit Codes
| Code | Meaning |
|---|---|
| 0 | Success |
| 1 | General error |
| 2 | Prerequisites not met |
| 3 | Network/API error |
| 4 | Git error |
| 5 | User cancelled |
| 6 | Rollback required (automatic) |
Limitations
- Use this skill only when the task clearly matches the scope described above.
- Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
- Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.
04 同仓库更多 Skills
vercel-sandbox
Run agent-browser + Chrome inside Vercel Sandbox microVMs for browser automation from any Vercel-deployed app. Use when the user needs browser automation in a Vercel app (Next.js, SvelteKit, Nuxt, Remix, Astro, etc.), wants to run headless Chrome without binary size limits, needs persistent browser sessions across commands, or wants ephemeral isolated browser environments. Triggers include "Vercel Sandbox browser", "microVM Chrome", "agent-browser in sandbox", "browser automation on Vercel", or any task requiring Chrome in a Vercel Sandbox.
slack
Interact with Slack workspaces using browser automation. Use when the user needs to check unread channels, navigate Slack, send messages, extract data, find information, search conversations, or automate any Slack task. Triggers include "check my Slack", "what channels have unreads", "send a message to", "search Slack for", "extract from Slack", "find who said", or any task requiring programmatic Slack interaction.
electron
Automate Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify, etc.) using agent-browser via Chrome DevTools Protocol. Use when the user needs to interact with an Electron app, automate a desktop app, connect to a running app, control a native app, or test an Electron application. Triggers include "automate Slack app", "control VS Code", "interact with Discord app", "test this Electron app", "connect to desktop app", or any task requiring automation of a native Electron application.
dogfood
Systematically explore and test a web application to find bugs, UX issues, and other problems. Use when asked to "dogfood", "QA", "exploratory test", "find issues", "bug hunt", "test this app/site/platform", or review the quality of a web application. Produces a structured report with full reproduction evidence -- step-by-step screenshots, repro videos, and detailed repro steps for every issue -- so findings can be handed directly to the responsible teams.
agent-browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.
shadcn
Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI, including chat interfaces. Provides project context, component docs, and usage examples. Applies when working with shadcn/ui, component registries, presets, --preset codes, or any project with a components.json file. Also triggers for "shadcn init", "create an app with --preset", "switch to --preset", and for building a design system or applying a DESIGN.md or brand spec to shadcn/ui.