By lswank
86 business AI specialists across 9 domains, 9 prompt-engineering frameworks, an on-demand prompt generator, and a meta-specialist for designing new specialists. For Claude Code (full plugin) or Claude Desktop (via the hosted MCP server). Clean-room reimplementation of the Black Book concept.
Structure a request using the A-I-M framework (Action, Intent, Metric).
Structure a request using the B-A-B framework (Before, After, Bridge).
Structure a request using the C-A-R-E framework (Context, Action, Result, Example).
You are the **Black Book Doctor**. Print a clean summary of what's installed in this plugin so the user can verify everything looks right.
Find the right Black Book specialist for a task you're trying to do.
Use for paid ad copy — Meta (Facebook/Instagram), Google Ads (search and display), LinkedIn, X, TikTok, YouTube ads. Produces multiple variants per ad set, not creative direction.
Use for all-hands meeting talking points, prepared remarks, monthly company-update scripts. Produces what the leader actually says, not the slide deck (use demo-deck-writer for slides).
Use for API and SDK documentation — endpoint references, getting-started guides, authentication docs, error-code tables, webhook docs, SDK quickstarts. Developer-audience. Pair with release-notes-writer for API change announcements.
Use for sales battlecards — competitor-by-competitor or category cards that arm reps in live deals. Produces field-ready cards, not market analyses (use competitive-analyst for analysis).
Use for long-form blog posts — guides, tutorials, opinion pieces, comparison posts, listicles, case studies. Produces full drafts with H2/H3 structure. Use seo-meta-writer separately for meta tags.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A Claude Code plugin: 86 role-built business specialists, the 9 prompt-engineering frameworks they're built on, and a generator that produces tailored prompts on demand.
A clean-room reimplementation of the "Black Book" concept — built independently from the public feature list, sharing none of its code or copy.
Landing page: https://lswank.github.io/black-book/
Quickstart: QUICKSTART.md — first 5 minutes
Examples: EXAMPLES.md — worked walkthroughs of common chains
FAQ: FAQ.md — common questions
L-E-D guide: docs/L-E-D.md — why the calibration framework has no specialist
Contributing: CONTRIBUTING.md
black-book.mcpb (~2.6 KB)The bundle proxies stdio MCP traffic to the hosted endpoint. Free tier rate limit: 30 requests per IP per hour. Self-host (below) for unlimited use.
For developers who'd rather use a script. Patches claude_desktop_config.json in place, preserving any existing entries:
macOS / Linux:
curl -fsSL https://black-book-host.vercel.app/install | bash
Windows (PowerShell):
iwr -useb https://black-book-host.vercel.app/install.ps1 | iex
Restart Claude Desktop after either.
Paste this into claude_desktop_config.json (~/Library/Application Support/Claude/ on macOS, %APPDATA%\Claude\ on Windows):
{
"mcpServers": {
"black-book": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://black-book-host.vercel.app/mcp"]
}
}
}
Restart Claude Desktop. The black-book tools (list_specialists, find_specialist, get_specialist, prompt_bank, flow, …) are now available.
Run the MCP server locally — unlimited use, fully private:
git clone https://github.com/lswank/black-book ~/code/black-book
cd ~/code/black-book/mcp && npm install
Then in claude_desktop_config.json:
{
"mcpServers": {
"black-book": {
"command": "node",
"args": ["/Users/YOUR_USERNAME/code/black-book/mcp/server.mjs"]
}
}
}
Restart Claude Desktop. See mcp/README.md for full details.
For the full experience — subagents, skills, slash commands, /flow, /prompt-bank, /list, /find, /doctor:
/plugin install https://github.com/lswank/black-book
Or clone into your Claude Code plugin directory:
git clone https://github.com/lswank/black-book ~/.claude/plugins/black-book
Either the hosted endpoint or the self-host pattern works — Black Book is a generic stdio / streamable HTTP MCP server. Point your client at https://black-book-host.vercel.app/mcp (HTTP) or node /path/to/black-book/mcp/server.mjs (stdio).
Each framework is a slash command. Pipe in a free-form description; get back a well-structured prompt.
| Command | Framework | Best for |
|---|---|---|
/tag | Task, Action, Goal | Quick ad-hoc requests |
/bab | Before, After, Bridge | Cold outreach, problem/solution copy |
/rtf | Role, Task, Format | Short structured outputs |
/care | Context, Action, Result, Example | Grounded, context-heavy work |
/rise | Role, Input, Steps, Expectation | Multi-step long outputs |
/aim | Action, Intent, Metric | Strategy, goal-setting |
/gro | Goal, Reason, Output | Lightweight planning, briefs |
/fit | Format, Input, Task | Transforms, restructuring |
/led | Level, Expectation, Direction | Calibrating tone and complexity |
Each framework asks for missing inputs and emits a copy-pastable, fenced prompt.
/prompt-bank — prompts on demandStatic prompt libraries are a 2023 artifact. Instead, /prompt-bank <task> does this:
npx claudepluginhub lswank/black-book --plugin black-bookDynamic multi-agent workflows for Grok Code — fan-out, pipelines, tournaments, adversarial verification, and loop-until-done harnesses, exposed as grok slash-commands. Each skill self-locates its bundled harness, so the plugin works in any directory (git repo or not) with no per-repo install.
Comprehensive PR review agents specializing in comments, tests, error handling, type design, code quality, and code simplification
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
Comprehensive startup business analysis with market sizing (TAM/SAM/SOM), financial modeling, team planning, and strategic research
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.
Complete creative writing suite with 10 specialized agents covering the full writing process: research gathering, character development, story architecture, world-building, dialogue coaching, editing/review, outlining, content strategy, believability auditing, and prose style/voice analysis. Includes genre-specific guides, templates, and quality checklists.