testing
Use when writing test specs for NL artifacts, running /nlpm:test, or setting up TDD workflows for skills, agents, commands, rules, hooks, and prompts.
- License 许可
- MIT
- 最近更新
- 2026-11-01
- Stars
- ★ 3
- Forks
- 1
01 前言 Front Matter
- name
- testing
- version
- 0.1.0
- description
- Use when writing test specs for NL artifacts, running /nlpm:test, or setting up TDD workflows for skills, agents, commands, rules, hooks, and prompts.
03 SKILL.md 正文
The NL-TDD Cycle
1. Write spec (.nlpm-test/artifact-name.spec.md) — define expectations
2. /nlpm:test — RED: spec fails (artifact doesn't exist)
3. Write the artifact — create the NL artifact
4. /nlpm:test — check if it passes
5. /nlpm:score — check quality score
6. Iterate until GREEN: all specs pass + score ≥ threshold
Spec File Format
Location: .nlpm-test/ directory in the project root (or alongside the artifact).
Filename convention: <artifact-name>.spec.md — matches the artifact filename without path.
---
artifact: agents/my-agent.md # path to the artifact being tested
type: agent # agent | skill | command | rule | hook | prompt
min_score: 85 # minimum /nlpm:score threshold for this artifact
---
Body sections (all optional — include what matters for this artifact):
triggers_on (skills and agents)
## Triggers On
Queries that SHOULD trigger this artifact:
- "review my database migrations before deploying"
- "check if these schema changes are safe"
- "audit the migration for breaking changes"
does_not_trigger_on (skills and agents)
## Does Not Trigger On
Queries that should NOT trigger this artifact:
- "write a migration for adding a users table"
- "help me with CSS styling"
- "deploy to production"
output_contains (agents and commands)
## Output Contains
Expected elements in the output:
- "## Migration Review" (heading present)
- "| Table | Change | Risk |" (table structure)
- severity classification (CRITICAL/HIGH/MEDIUM/LOW)
output_format (agents and commands)
## Output Format
The output should be a markdown report with:
1. Summary section with counts
2. Findings table with columns: File, Issue, Severity
3. Action items list
handles_input (commands)
## Handles Input
| Input | Expected Behavior |
|-------|------------------|
| (empty) | Score all artifacts in cwd |
| directory path | Score artifacts in that directory |
| nonexistent path | Error: "Directory not found: {path}" |
| file path | Score that single file |
follows_rules (rules)
## Follows Rules
Code that SHOULD comply:
```python
result: Result[User, AppError] = get_user(id)
Code that SHOULD violate:
user = get_user(id).unwrap() # should be flagged
### frontmatter_valid (all types)
```markdown
## Frontmatter Valid
Required fields:
- description: present and trigger-style ("Use when...")
- model: sonnet
- tools: [Read, Glob, Grep]
- skills: [nlpm:conventions, nlpm:scoring]
Test Results Format
NLPM Test Report
Spec Artifact Result Details
─────────────────────────────────────────────────────────────────────────────────
my-agent.spec.md agents/my-agent.md PASS 5/5 checks
my-skill.spec.md skills/core/SKILL.md FAIL 3/5 checks
✗ Trigger: "optimize React hooks" → predicted NO trigger (expected YES)
✗ Score: 68/100 (min: 85)
Overall: 1 passed, 1 failed (50%)
RED items (fix these):
1. skills/core/SKILL.md — trigger gap: "optimize React hooks" not covered by description
2. skills/core/SKILL.md — score 68 < min 85: missing <example> blocks (-15)
Best Practices for Specs
- Write specs BEFORE writing the artifact (TDD discipline)
- 5-10 trigger queries for skills/agents (mix positive and negative)
- Include edge cases in
handles_inputfor commands min_scoreshould match your project's threshold (default 85 for new artifacts, 70 for legacy)- Specs are living documents — update when behavior requirements change
Spec File Discovery
The tester discovers specs by:
- Looking in
.nlpm-test/directory - Matching spec filename to artifact filename:
my-agent.spec.md→agents/my-agent.md(uses theartifact:frontmatter field) - If
artifactpath doesn't exist → spec is RED by default (artifact not yet created — this is the TDD "write test first" state)
04 同仓库更多 Skills
seedance-prompt-en
Write effective prompts for Jimeng Seedance 2.0 multimodal AI video generation. Use when users want to create video prompts using text, images, videos, and audio inputs with the @ reference system. Covers camera movements, effects replication, video extension, editing, music beat-matching, e-commerce ads, short dramas, and educational content.
onepage-pdf
Convert an HTML page into a single continuous-page PDF (one tall page, no pagination breaks), preserving desktop layout, backgrounds and selectable text. Use when the user asks to turn an HTML file/report/proposal into a "single-page PDF", "long PDF", "不分页 PDF", "单页 PDF", "长图式 PDF", or complains that an HTML-to-PDF export breaks into pages or cuts content. Supports optional string redaction with leak verification.
native-feel-cross-platform-desktop
Use when the user is designing, prototyping, or rewriting a desktop app that must run on multiple OSes (macOS + Windows, optionally Linux) AND feel indistinguishable from a native app to its users — fast launch, native windowing, native input handling, native materials. Trigger words include "cross-platform desktop", "Electron alternative", "Tauri vs native", "WebView wrapper", "near-native performance", "Raycast architecture", "WebKit/WebView2 quirks", "WKWebView", "system tray app", "global hotkey app", "launcher app". Do NOT trigger this skill for pure web apps, pure mobile apps, or for greenfield projects that have no native-feel requirement.
gstack
Router for the gstack skill suite. (gstack)
fireworks-tech-graph
Use when the user wants to create any technical diagram - architecture, data flow, flowchart, sequence, agent/memory, or concept map - and export as SVG+PNG. Trigger on: "画图" "帮我画" "生成图" "做个图" "架构图" "流程图" "可视化一下" "出图" "generate diagram" "draw diagram" "visualize" or any system/flow description the user wants illustrated.
dbs-video-extract
接收抖音、小红书、微信视频号等短视频链接或分享文案,查询作品/账号数据并提取语音文字稿,按作者和标题归档为 Markdown。用户希望完整解析短视频、取得数据或文字稿时使用;运行前检查 TikHub 与轻抖使用凭证,缺少时用新手能理解的方式引导购买和安全配置。