{"id":37372323,"url":"https://github.com/joseph0926/prompt-smith","last_synced_at":"2026-01-20T09:06:17.580Z","repository":{"id":331530470,"uuid":"1126459035","full_name":"joseph0926/prompt-smith","owner":"joseph0926","description":"prompt quality management for Claude Code","archived":false,"fork":false,"pushed_at":"2026-01-10T10:34:19.000Z","size":672,"stargazers_count":4,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-01-10T22:19:26.155Z","etag":null,"topics":["claude-code","claude-plugin","claude-skills"],"latest_commit_sha":null,"homepage":"","language":"JavaScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/joseph0926.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","threat_model":null,"audit":null,"citation":null,"codeowners":"CODEOWNERS","security":"SECURITY.md","support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-01-02T00:35:39.000Z","updated_at":"2026-01-10T10:34:15.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/joseph0926/prompt-smith","commit_stats":null,"previous_names":["joseph0926/prompt-smith"],"tags_count":10,"template":false,"template_full_name":null,"purl":"pkg:github/joseph0926/prompt-smith","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/joseph0926%2Fprompt-smith","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/joseph0926%2Fprompt-smith/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/joseph0926%2Fprompt-smith/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/joseph0926%2Fprompt-smith/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/joseph0926","download_url":"https://codeload.github.com/joseph0926/prompt-smith/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/joseph0926%2Fprompt-smith/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":28477210,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-01-16T03:13:13.607Z","status":"ssl_error","status_checked_at":"2026-01-16T03:11:47.863Z","response_time":107,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["claude-code","claude-plugin","claude-skills"],"created_at":"2026-01-16T05:00:30.445Z","updated_at":"2026-01-16T05:00:53.337Z","avatar_url":"https://github.com/joseph0926.png","language":"JavaScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# prompt-smith\n\n\u003e Prompt quality management for Claude Code\n\n\u003e **[한국어 문서 (Korean)](README.ko.md)**\n\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n[![Version](https://img.shields.io/badge/version-3.3.1-blue.svg)](https://github.com/joseph0926/prompt-smith/releases)\n\n![ps_demo_02](https://github.com/user-attachments/assets/82cf7bf6-433a-4f3b-be3d-7389e341afaf)\n\n**Core Features**:\n\n- 8-Point Quality Check (LINT / BUILD / Review / Intercept)\n- Eval Mode: A/B testing with test datasets\n- Long Context: 200K token optimization\n- Token Management: Cost estimation \u0026 context validation\n- Structured Outputs: 100% JSON validity\n\n## When to Use Which Mode?\n\n| Situation                     | Recommended | Command                                                          |\n| ----------------------------- | ----------- | ---------------------------------------------------------------- |\n| \"Is this prompt good enough?\" | LINT        | `/ps:lint \u003cprompt\u003e`                                              |\n| \"Make this prompt better\"     | Review      | `/ps:r \u003cprompt\u003e`                                                 |\n| \"Just fix it and run\"         | Intercept   | `/ps:a \u003cprompt\u003e`                                                 |\n| \"Design a new prompt\"         | BUILD       | `/ps:build \u003cgoal\u003e`                                               |\n| \"Compare prompt variants\"     | Eval        | See [Eval Mode](skills/prompt-smith/playbooks/eval/eval-mode.md) |\n\n## Quick Start\n\n### Commands at a Glance\n\n| Command             | Description                                               | Best For          |\n| ------------------- | --------------------------------------------------------- | ----------------- |\n| `/ps:r \u003cprompt\u003e`    | **Review Mode** - Show improvements, await approval       | Daily use (safe)  |\n| `/ps:lint \u003cprompt\u003e` | **LINT Mode** - Full diagnosis with test cases            | Debugging prompts |\n| `/ps:a \u003cprompt\u003e`    | **Intercept Mode** - Auto-improve and execute immediately | Quick tasks       |\n| `/ps:build \u003cgoal\u003e`  | **BUILD Mode** - Design from requirements                 | New prompts       |\n| `/ps:help [topic]`  | **Help** - Show usage guide and command reference         | Getting started   |\n\n**Quick Example**:\n\n```\n/ps:r Write a function to parse JSON\n```\n\nShows score (3/10 → 8/10), changes, asks \"Proceed? (y/n/e)\"\n\n---\n\n### Option 1: Plugin Install (slash commands enabled)\n\n1. In VS Code, type `/plugin` to open plugin terminal\n2. Press `Tab` to navigate to \"Add Marketplace\"\n3. Enter: `joseph0926/prompt-smith`\n4. Press `Tab` to navigate to \"Install Plugin\"\n5. Select `ps@prompt-smith`\n\n![Install Guide](assets/install_g_01.png)\n\n### Option 2: Local Plugin (development)\n\n```bash\ngit clone https://github.com/joseph0926/prompt-smith\nclaude --plugin-dir ./prompt-smith\n```\n\n### Option 3: Skill Only (natural language triggers)\n\n```bash\ngit clone https://github.com/joseph0926/prompt-smith\ncp -r prompt-smith/skills/prompt-smith ~/.claude/skills/\n```\n\n\u003e **Note**: This method only enables natural language triggers (`use prompt-smith -r`), not slash commands.\n\n### Usage\n\n**Slash Commands (recommended)**:\n\n```\n/ps:r your prompt here            # Review Mode (single line)\n/ps:r your prompt here\nwith multiple lines               # Review Mode (multiline)\n/ps:a your prompt here            # Intercept Mode\n/ps:lint your prompt              # LINT Mode\n/ps:build requirements            # BUILD Mode\n```\n\n\u003e **Note**: Triple backticks (```) are optional. Plain text and multiline input are fully supported.\n\n**Natural Language (alternative)**:\n\n```\nuse prompt-smith -r Your prompt here\n```\n\n## Real Examples\n\n### LINT Mode (Diagnose)\n\n```\n/ps:lint Analyze user feedback\n```\n\n**Output**: Score 3/10 → Top 3 issues → Improved prompt → 5 test cases\n\n### Review Mode (Safe Improvement)\n\n```\n/ps:r Write a function to parse JSON\n```\n\n**Output**:\n\n```\n+----------------------------------------------------------+\n| Score: 3/10 -\u003e 8/10 (+5)                                 |\n+----------------------------------------------------------+\nChanges:\n- [+] Added: Software engineer role\n- [+] Added: Error handling examples\n- [~] Modified: Specified return format\n\nProceed? (y/n/e)\n```\n\n### Intercept Mode (Auto-fix)\n\n```\n/ps:a Summarize this document\n```\n\n**Output**: Auto-improved if +2 points possible → Executes immediately\n\n## What's New in v3.x\n\n### v3.3.1 (Latest)\n\n| Feature                   | Description                                                         |\n| ------------------------- | ------------------------------------------------------------------- |\n| **Examples \u0026 TOCs**       | Expanded examples and added TOCs to long references                 |\n| **Structured Outputs**    | Stronger guidance for schema mismatch and JSON mode caution         |\n\n### v3.3.0\n\n| Feature                   | Description                                          |\n| ------------------------- | ---------------------------------------------------- |\n| **ps:a Token Efficiency** | Minimal output: `[Prompt Smith] 활성화됨 (X→Y점)`       |\n| **ps:r Intent Capture**   | AskUserQuestion for format/detail/constraints/goals  |\n\n### v3.2.0\n\n| Feature                | Description                                   |\n| ---------------------- | --------------------------------------------- |\n| **Eval Mode**          | A/B prompt comparison with test datasets      |\n| **Long Context**       | 200K token optimization (chunking, placement) |\n| **Token Management**   | Cost estimation, context window validation    |\n| **Structured Outputs** | 100% JSON validity techniques                 |\n\n### v3.0.0 - v3.1.0\n\n| Feature                 | Description                                        |\n| ----------------------- | -------------------------------------------------- |\n| **Progressive Loading** | Skills load on-demand                              |\n| **Hooks**               | Auto-LINT on prompt submit (optional)              |\n| **Agents**              | 3 specialized: optimizer, reviewer, test-generator |\n| **GitHub Actions**      | CI/CD integration                                  |\n\nSee [CHANGELOG.md](CHANGELOG.md) for full release notes.\n\n## Modes\n\n| Mode      | Slash Command       | Natural Language        | Description                       |\n| --------- | ------------------- | ----------------------- | --------------------------------- |\n| Review    | `/ps:r \u003cprompt\u003e`    | `use prompt-smith -r`   | Show improvements, await approval |\n| Intercept | `/ps:a \u003cprompt\u003e`    | `use prompt-smith -a`   | Auto-improve and execute          |\n| LINT      | `/ps:lint \u003cprompt\u003e` | `lint this prompt`      | Diagnose existing prompts         |\n| BUILD     | `/ps:build \u003cgoal\u003e`  | `build a prompt for...` | Design from requirements          |\n\n## 8-Point Quality Check\n\n| #   | Dimension        | Score | Note         |\n| --- | ---------------- | ----- | ------------ |\n| 1   | ROLE             | 0-2   | Base         |\n| 2   | CONTEXT          | 0-2   | Base         |\n| 3   | INSTRUCTION      | 0-2   | Base         |\n| 4   | EXAMPLE          | 0-2   | Base         |\n| 5   | FORMAT           | 0-2   | Base         |\n| 6   | SUCCESS_CRITERIA | 0-2   | Base (v2.7+) |\n| 7   | STATE_TRACKING   | 0-2   | Extended     |\n| 8   | TOOL_USAGE       | 0-2   | Extended     |\n\n**Base Score**: Dimensions 1-6 (max 12 → normalized to 10)\n**Extended Score**: All 8 dimensions (normalized)\n\n## Documentation\n\n### Getting Started\n\n- [Quick Start](skills/prompt-smith/onboarding/quick-start.md)\n- [First LINT](skills/prompt-smith/onboarding/first-lint.md)\n- [First BUILD](skills/prompt-smith/onboarding/first-build.md)\n\n### Mode Guides\n\n- [LINT Mode](skills/prompt-smith/playbooks/lint/full-lint.md)\n- [BUILD Mode](skills/prompt-smith/playbooks/build/build-mode.md)\n- [Review Mode](skills/prompt-smith/playbooks/intercept/review-mode.md)\n- [Intercept Mode](skills/prompt-smith/playbooks/intercept/intercept-mode.md)\n- [Eval Mode](skills/prompt-smith/playbooks/eval/eval-mode.md)\n\n### Advanced (v3.2.0)\n\n- [Long Context Optimization](skills/prompt-smith/references/long-context-optimization.md)\n- [Token Management](skills/prompt-smith/references/token-management.md)\n- [Structured Outputs](skills/prompt-smith/references/structured-outputs.md)\n- [Prompt Chaining](skills/prompt-smith/playbooks/prompt-chaining.md)\n\n### Reference\n\n- [8-Point Quality Checklist](skills/prompt-smith/references/quality-checklist.md)\n- [Anti-Patterns](skills/prompt-smith/references/anti-patterns.md)\n- [Technique Priority](skills/prompt-smith/references/technique-priority.md) (Anthropic recommended)\n- [Claude 4.x Best Practices](skills/prompt-smith/references/claude-4x-best-practices.md)\n\n## Troubleshooting\n\n### Slash commands not working\n\n1. Verify plugin: In VS Code, type `/plugin` → check installed list\n2. Restart VS Code or reload Claude Code extension\n3. Try reinstalling: Remove plugin, then reinstall\n\n### Score seems wrong\n\n- Extended items (STATE_TRACKING, TOOL_USAGE) only apply to multi-step or tool-using prompts\n- If N/A, they're excluded from score calculation\n\n## Platform\n\n- **Claude Code** (CLI / VS Code extension)\n\n## Contributing\n\nWe welcome contributions! See [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.\n\n- [Report a bug](https://github.com/joseph0926/prompt-smith/issues)\n- [Security issues](SECURITY.md)\n\n## License\n\nMIT License - see [LICENSE](LICENSE)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjoseph0926%2Fprompt-smith","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjoseph0926%2Fprompt-smith","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjoseph0926%2Fprompt-smith/lists"}