Skip to content
Agentmart
Submit a tool

validate

Run validation tests for your current chapter implementation.

Skill by tbhbooks

IndexedMIT
Write a review

Install

Works with Claude Code. Read the repository before you give any tool access to your data.

npx agentmarket install skill/tbhbooks-validate

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Run validation tests for your current chapter implementation. Fetches tests from GitHub and checks your code against the spec's test suite. Usage - /tbh:validate or /tbh:validate 3

Use cases: testing

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/validate

Tools it pre-approves 6

From allowed-tools in SKILL.md: the agent may use these without asking you each time.

  • Bash
  • Read
  • Grep
  • Glob
  • WebFetch
  • Write

What's inside

Skill name
validate
Description
Run validation tests for your current chapter implementation. Fetches tests from GitHub and checks your code against the spec's test suite. Usage - /tbh:validate or /tbh:validate 3
License
MIT
Allowed tools
Bash, Read, Grep, Glob, WebFetch, Write
Instructions
91 lines in SKILL.md · view on GitHub

Before installing

  • Pre-approves any shell command
  • Instructions call the network

Files (1)

  • SKILL.md

SKILL.md (first lines)

# TBH Validate

Run the chapter's validation tests against the reader's implementation.

## Fetch Preamble

1. **Read `.tbh/config.json`** for active book, chapter, `raw_base`, and `spec_path`. If missing, tell reader to run `/tbh:setup` first.
2. **Determine chapter number.** Use argument if provided, otherwise `current_chapter`. Format as two digits.
3. **Fetch the test file from GitHub:**
   ```bash
   mkdir -p .tbh/cache/validation
   curl -sf "<raw_base>/<spec_path>/ch<NN>/validation/test_ch<NN>.py" -o .tbh/cache/validation/test_ch<NN>.py
   ```
   If fetch fails, use cached version. If no cache, tell reader: "No validation tests available for this chapter yet."

## Steps

4. **Read the fetched test file** to understand what it expects (command to run, expected output format).

5. **Run the tests:**
   ```bash
   cd <reader's project directory>
   python -m pytest .tbh/cache/validation/test_ch<NN>.py -v 2>&1 || python .tbh/cache/validation/test_ch<NN>.py 2>&1
   ```

6. **Report results clearly:**

   If all pass:
   ```
   ✓ Chapter <N> — All tests passing!

   Run /tbh:next to see what's ahead.
   ```
   Update `chapters_completed` in config and progress.

   If some fail:
   ```
   Chapter <N> Validation:
     ✓ test_tool_interface_exists
     ✓ test_read_file_tool
     ✗ test_search_code_tool — pattern parameter missing
     ✗ test_tool_registry_match — returned 0 results

   2/4 passing

   Hint: Read the interface-spec for search_code's parameter schema.
   Run /tbh:hint for more guidance.
   ```

7. **Never reveal test implementation details** that would give away the answer. Summarize what failed, not how to fix it.

## Star Prompt (Chapter 1 completion, per book)

After Chapter 1 passes all tests, check `.tbh/progress.json` for a per-book star tracking field.

The `progress.json` structure for this:
```json
{
  "book": "agents",
  "started_at": "...",
  "chapters": {},
  "stars_prompted": {
    "agents": true,
    "llm-inference": true
  }
}
```

**Logic:** Read the active book ID from config. Check if `stars_prompted.<active_book>` is `true` in `progress.json`.

If it is NOT `true` (or `stars_prompted` doesn't exist yet):

1. **Show a message after the success report:**
   ```
   ✓ Chapter 1 complete — you just built your first agent loop!

   tbh, if this is helping you learn, a star goes a long way:
   👉 <repo URL from config>

   Or if you have the GitHub CLI:
     gh repo star <org/repo-name>
   ```

2. **Update `.tbh/progress.json`** — set `stars_prompted.<active_book>` to `true`.

If already `true` for this book, skip entirely. This means:
- Each book gets one star prompt (after its Ch 1)
- Switching to a new book and completing its Ch 1 triggers a fresh prompt for that book's repo
- No book ever prompts twice

From tbhbooks/tbh-skill, MIT.

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Apr 15, 2026 (6 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from tbhbooks

Cite this page

validate skill. Agent Mart. https://agentmart.dev/skills/tbhbooks-validate (last updated 2026-10-24)