LIVE MARKETPLACE

The Universal
Agent Marketplace

Every package manager. Every registry. Every tool. One unified platform that resolves, installs, analyzes, and exposes CLI tools as AI-native skills.

0
Installed Tools
0
Generated Skills
0
MCP Tools
0
Source Registries
agents-cli
$ agents-cli search "AI workflow automation"
0
Tools Installed
0
Skills Generated
0
MCP Tools
0
Domain Plugins
0
Registries

Trusted by developers building the agentic future

Five Steps to Agent-Ready

Every tool follows the same battle-tested pipeline — from raw source to AI-native skill.

🔍
Resolve
Detect source format, fetch metadata, discover versions across 7 registries
agents-cli add pypi:ruff
📥
Install
Download, extract, build with branch fallback (main → master → develop)
pip install ruff → ~/.agents-cli/
🧪
Analyze
Deep --help parsing, recursive probing, schema introspection, smoke test
deepProbe(bin, {maxDepth: 5})
Generate
SKILL.md with frontmatter, refs, scripts. Quality gates: trigger ≥ 0.80
generateSkillDirectory(tool)
🚀
Expose
MCP bridge server, Claude Code native integration, agent-ready API
agents-cli mcp start
MCP v1.0 Claude Code

Every Source. One Command.

Agents-CLI unifies the fragmented tool ecosystem. Install from any registry with a single, consistent interface. No more juggling package managers.

🐙

GitHub

owner/repo

Direct repository access. Clones, builds, and analyzes any GitHub project. Supports branch fallback (main → master → develop), release detection, and automatic binary inference.

Auto-build Release tracking Binary inference README extraction
agents-cli add anthropics/claude-code
⭐ Most versatile 200+ tools
📦

npm

npm: / @scope/

Node.js ecosystem. CLI tools, frameworks, and libraries. Scoped packages supported natively. Automatic dependency resolution and global binary detection.

Scoped packages npx support Version pinning Lockfile integrity
agents-cli add npm:eslint agents-cli add @anthropic-ai/sdk
🚀 Fastest installs 2.5M+ packages
🐍

PyPI

pypi:

Python Package Index. AI/ML tools, data science utilities, automation agents. Virtual environment isolation. Bare names auto-resolve to PyPI.

venv isolation AI/ML focused Auto-resolve bare names Entry point detection
agents-cli add pypi:ruff agents-cli add httpie
🧠 AI/ML leader 500K+ packages
🦀

crates.io

crates:

Rust ecosystem. High-performance CLI binaries with zero-cost abstractions. Cargo.toml [[bin]] parsing for multi-binary crates. Native compilation.

Native binaries Multi-binary support Cargo.toml parsing Zero-cost performance
agents-cli add crates:ripgrep
⚡ Fastest runtime 150K+ crates
PREMIUM
🤖

CLI-Anything

cli-anything:

Revolutionary. Wraps any GUI or desktop application with AI-generated CLI bindings. Blender, Photoshop, Figma, VS Code — turn any app into a scriptable agent tool.

GUI → CLI AI harness generation App profiling Any desktop app NEW
agents-cli add cli-anything:blender agents-cli add cli-anything:figma
💥 Game changer Unlimited apps
PREMIUM
🌐

MCP-to-CLI

mcp:

Bridge Model Context Protocol servers into CLI tools. Introspects MCP servers, extracts tool schemas, and generates agent-consumable skill definitions. The glue between MCP and everything else.

MCP introspection Schema extraction Auto skill gen Bidirectional bridge NEW
agents-cli add mcp:filesystem agents-cli add mcp:sqlite
🔗 MCP native 534 tools
📁

Local

./ or /

Your own tools. Point at any directory to analyze, generate skills, and expose to agents. Perfect for proprietary internal tools and private repositories.

Private tools Instant analysis No upload needed Enterprise ready
agents-cli add ./my-internal-tool
🔒 Air-gapped OK Any directory

Unified Pipeline — Every Source, Same Process

01
Resolve
Detect format, fetch metadata, version detection
02
Install
Download, extract, build with branch fallback
03
Analyze
--help parsing, deep probe, schema introspection
04
Generate
SKILL.md, refs, scripts, quality gates
05
Expose
MCP bridge, agent-ready, AI-native

Orchestrate Multi-Agent Pipelines

Workflows chain multiple skills into automated pipelines. Scout data, generate content, process media, deploy — all in one composable flow. AI-composed or hand-crafted.

🧠

AI Composition

Describe what you want in natural language. The agentic composer analyzes your installed skills, discovers compatible tools, and proposes an optimized workflow.

agents-cli compose "Python CI pipeline"
🔄

Iterative Refinement

3-5 iteration loops with validation. Each round improves step completeness, data flow validity, env var documentation, and setup runnability. Target quality ≥ 0.8.

--iterations 5 --sandbox --creative
🎯

4-Axis Quality Gates

Every workflow must pass: step completeness, data flow validity, env var documentation, and setup runnability. Each axis ≥ 0.5 to ship.

Step Completeness
0.92
Data Flow Validity
0.88
Env Documentation
0.95
Setup Runnability
0.85

Featured Workflows

🔍

Topological Sorting

Steps are ordered by dependency analysis. Cross-script imports and file I/O chains determine execution order automatically.

📊

Data Flow Inference

Outputs from one step automatically connect to inputs of the next. Artifact tracking across the entire pipeline.

🛡️

Sandbox Validation

Docker-based sandbox testing ensures workflows run correctly before deployment. Isolated execution environment.

⚙️

Env Var Merging

All environment variables across steps are collected, deduplicated, and documented. setup.sh validates every requirement.

⏱️

Duration Estimation

Intelligent estimation of total pipeline runtime based on step analysis. Plan resource allocation accurately.

📝

Script Generation

Auto-generates run.sh orchestrator, setup.sh validator, SKILL.md documentation, and workflow.md with pipeline diagrams.

Compose Workflows in Seconds

Natural language → validated pipeline. The composer discovers skills across all registries, infers data flow, and generates production-ready scripts.

--from-skills Compose from specific installed skills
--creative Use all 4 discovery methods for maximum coverage
--sandbox Validate in Docker before shipping
--domain Scope discovery to a specific domain
compose
$ agents-cli compose "content creation pipeline" --creative --sandbox
➜ Discovering skills across 7 registries...
➜ Found 12 compatible skills (semantic + domain + graph)
➜ Proposing workflow with TieredLLMClient...
✓ Iteration 1/5 — quality: 0.72
✓ Iteration 2/5 — quality: 0.81
✓ Iteration 3/5 — quality: 0.89 ✔️
➜ Running sandbox validation...
✓ All 4 quality axes passed (≥ 0.5)
➜ Generated: SKILL.md, run.sh, setup.sh, workflow.md
🎉 Workflow "content-creation-pipeline" ready!

Build Anything. Ship Everything.

From atomic skills to enterprise agent teams — every building block for the agentic future.

📜

Skills

Atomic agent capabilities. SKILL.md with frontmatter, references, scripts. Quality-gated with trigger score ≥ 0.80.

393 available
🧩

Plugins

Claude Code spec compliant. Multi-skill bundles with commands, agents, hooks, and settings. 52 domain plugins generated.

52 domains

Workflows

Multi-step pipelines. Topological ordering, data flow inference, env var merging. AI-composed or hand-crafted.

6 featured
🛡

Hook Bundles

Pre/post hooks for 7 event types. Pre-command, post-edit, pre-task, session start/end. Behavioral guardrails.

7 event types
🤖

Agent Definitions

Autonomous agent personas with system prompts, tool access, and capability boundaries. Markdown-based, composable.

Pro tier
👥

Agent Teams

Multi-agent orchestration. Coordinator + specialist agents with shared context, handoff protocols, and consensus.

Enterprise
💻

Harnesses

CLI-Anything generated wrappers. GUI apps exposed as command-line tools with full schema introspection.

Unlimited

Why Subscribe

Every feature is designed for agents-first development. Subscribe to unlock the full power of the platform.

🔥

Skill Forge

3-layer pipeline: structural → AI-enhanced → quality-gated. Generate production-grade skills from any tool with deep schema probing, README extraction, and auto-description.

  • Deep probe with configurable depth
  • Trigger score optimization (avg 1.000)
  • Quality score ≥ 6/10 guaranteed
  • Content score ≥ 5/10 guaranteed
  • Factory mode for 3-layer refinement
  • Batch processing with CSV export
Starter+
🕷️

Crawl System

Scale discovery across all registries. Adaptive concurrency, exponential backoff, install-analyze-prune cycle. Seed from PyPI, npm, crates, GitHub, and MCP registries.

  • Adaptive concurrency (auto CPU detection)
  • Token bucket rate limiting
  • Exponential backoff (1min → 30min)
  • Install-analyze-prune cycle
  • Cross-registry seeding
  • Libraries.io integration
Pro+
🌐

MCP Bridge

Expose every installed tool as an MCP-compatible server. Bidirectional: CLI → MCP and MCP → CLI. 534 tools available through the bridge.

  • Automatic tool exposure
  • Schema introspection
  • MCP v1.0 compatible
  • mcp2cli bidirectional bridge
  • 534+ tools available
  • Claude Code native integration
Starter+
🔍

Hybrid Search

FTS5 full-text search + sqlite-vec KNN embeddings. Search across skills, workflows, and tools with natural language queries. Domain-scoped filtering.

  • FTS5 full-text indexing
  • sqlite-vec embedding search
  • Natural language queries
  • Domain tree filtering
  • Paginated results
  • Cross-product-type search
Starter+
💰

Creator Economy

Publish skills and earn. 80% revenue share to creators. Per-call metering, monthly caps, and enterprise licensing. Your tools, your revenue.

  • 80/10/10 revenue split
  • Per-call metering
  • Monthly subscription caps
  • Creator dashboard
  • Earnings tracking
  • Marketplace publishing
Pro+

Try It Live

Experience the full power of Agents-CLI right in your browser. Type commands, explore sources, and inspect workflows.

💻 Live CLI Simulator

agents-cli simulator
Welcome to agents-cli simulator. Type 'help' for commands.
$

🔄 Source Adapter Picker

agents-cli add owner/repo

Clones any GitHub repository, detects language, builds from source, and analyzes all CLI entry points.

Clone
git clone
Build
npm/cargo/pip
Probe
--help parse
Generate
SKILL.md
Expose
MCP ready
agents-cli add npm:eslint

Installs from npm registry with full dependency resolution. Detects global binaries and CLI entry points.

Resolve
npm registry
Install
npm install
Analyze
bin detection
Generate
SKILL.md
Expose
MCP ready
agents-cli add pypi:ruff

Creates isolated venv, installs via pip, detects entry points. Bare names auto-resolve to PyPI.

Resolve
PyPI API
Venv
pip install
Detect
entry points
Generate
SKILL.md
Expose
MCP ready
agents-cli add crates:ripgrep

Compiles Rust crates natively. Parses Cargo.toml for multi-binary support. Zero-cost performance.

Resolve
crates.io
Build
cargo build
Parse
Cargo.toml
Generate
SKILL.md
Expose
MCP ready
agents-cli add cli-anything:blender

Wraps ANY desktop application with AI-generated CLI bindings. GUI → CLI transformation.

Profile
app analysis
Harness
AI bindings
Test
smoke test
Generate
SKILL.md
Expose
MCP ready
agents-cli add mcp:filesystem

Introspects MCP servers, extracts tool schemas, generates CLI wrappers. Bidirectional bridge.

Connect
MCP server
Introspect
tool schemas
Wrap
CLI binding
Generate
SKILL.md
Expose
MCP ready
agents-cli add ./my-internal-tool

Analyze any local directory. No upload needed. Perfect for private tools and enterprise repos.

Scan
local dir
Detect
entry points
Analyze
--help parse
Generate
SKILL.md
Expose
MCP ready

⚡ Workflow Inspector

Content Creation Pipeline — Hover to inspect
Tool: web-scout
Command: scout --trending --limit 10
Env: SCOUT_API_KEY
🔍 Scout web-scout
Tool: ghostwriter
Command: write --tone professional
Env: OPENAI_API_KEY
✍️ Write ghostwriter
Tool: dalle-cli
Command: generate --style editorial
Env: OPENAI_API_KEY
🖼️ Image dalle-cli
Tool: cms-push
Command: publish --platform wordpress
Env: WP_API_TOKEN
📤 Publish cms-push

Explore the Catalog

Loading...
Loading more...

Want more? Subscribe to unlock unlimited installs, premium sources, and workflow execution.

Choose Your Plan

Start free. Scale as you build. Every plan includes the core marketplace.

Free

$ 0 /forever
  • 3 installs/month
  • GitHub, npm, PyPI, crates, local sources
  • Basic skill generation
  • MCP bridge (read-only)
  • Community support
  • CLI-Anything
  • MCP-to-CLI
  • Workflow execution
  • Intelligence layer
  • Creator economy
Current Plan

Starter

$29 $ 14.99 /mo
  • 50 installs/month
  • All 7 source registries
  • CLI-Anything adapter
  • MCP-to-CLI bridge
  • 5 workflow installs/month
  • Skill Forge (structural)
  • MCP bridge (full)
  • Hybrid search
  • Intelligence layer
  • Publish to marketplace

Enterprise

$199 $ 99 /mo
  • Unlimited everything
  • Everything in Pro
  • Agent Teams (multi-agent)
  • Private registry
  • Air-gapped deployment
  • Custom domain plugins
  • SLA + dedicated support
  • SSO / SAML
  • Audit logs
  • Custom integrations

Feature Comparison

Feature Free Starter Pro Enterprise
Monthly installs350500
GitHub source
npm source
PyPI source
crates.io source
Local source
CLI-Anything
MCP-to-CLI bridge
Workflow viewView only5/mo
Workflow compose
Workflow publish
Skill ForgeBasicStructural3-layer + AI3-layer + AI
Intelligence layer
Crawl system
Hybrid searchFTS only
MCP bridgeRead-onlyFullFullFull
Publish to marketplace
Creator revenue (80%)
Agent definitions
Agent teams
Private registry
SSO / SAML
SupportCommunityEmailPriorityDedicated + SLA

Ready to build the future?

Join thousands of developers using Agents-CLI to bridge the gap between tools and AI agents.

Explore Free Plan
🔒 Secure checkout via Stripe 💳 Cancel anytime ⚡ Instant access