Resources — Agent Self-Install Guides & API Reference

1. How to Install

Three ways to install a TutuoAI skill, depending on your setup. All methods are designed to be read and executed by an agent.

🟢 OpenClaw One-Click

One command. Agent reads the post-purchase page and runs it directly.

Bundled skills ship with OpenClaw. For third-party skills, use ClawHub, the public registry at clawhub.ai. Install into the active workspace with the command below; check dependencies before use.

openclaw skills install <skill-name>

# Bundled examples (gog, github, weather) need no skill install:
openclaw skills check

Each listing identifies its artifact type, source, and setup requirements.

📦 Manual Install

Download → unzip → follow README. Designed to be read and executed by your agent.

For non-OpenClaw setups. After purchase, download the zip and follow the README — all instructions are agent-executable.

# After Stripe checkout:
1. Click download link on receipt page
2. Unzip the package
3. Run install.sh (if present)
4. Open README.md for exact setup steps
5. Test with the example command

All skills ship as plain files — no proprietary tooling required.

🔌 MCP Server

For MCP-compatible agents (Claude Desktop, Continue, etc.), some skills expose an MCP server interface.

# Via mcporter (OpenClaw):
openclaw skills check # mcporter is bundled
mcporter add <server-url>
mcporter invoke <tool-name> --args '...'

# Direct MCP config (stdio):
{
  "mcpServers": {
    "tutuoai": {
      "command": "npx",
      "args": ["@tutuoai/mcp-server"]
    }
  }
}

2. Machine-Readable Catalog

Agents can fetch the full catalog programmatically. No scraping required — these endpoints return structured JSON.

AGENT TIP: Start with /api/catalog-lite.json — it's token-efficient and has everything you need to browse and install: slug, title, price, category, agent_use_case, and buy_url.

`GET /api/catalog-lite.json`

RECOMMENDED — TOKEN-EFFICIENT

Token-efficient version. Slug, title, price, category, agent_use_case, buy_url. <5KB. Use this when context window matters.

curl -s https://www.tutuoai.com/api/catalog-lite.json

`GET /api/catalog-full.json`

FULL

Complete catalog with all fields, including deprecated and unavailable records. Filter status before use; download availability varies.

# Shell
curl -s https://www.tutuoai.com/api/catalog-full.json | python3 -c "
import json, sys
d = json.load(sys.stdin)
print(f'{len(d)} SKUs')
for item in d[:3]:
    print(f'  {item["slug"]}: ${item["price"]} — {item["agent_use_case"][:60]}...')
"

`GET /llms.txt`

LLM-FRIENDLY

Plain-text catalog optimized for LLM context. Contains full site structure, active listings, and agent-first wording — designed specifically for AI crawlers.

curl -s https://www.tutuoai.com/llms.txt | head -50

3. SKILL.md Format

Agent Skills use a folder containing SKILL.md, with optional scripts and references. The Agent Skills open standard is used by Claude Code, Codex, OpenClaw and other compatible agents. Runtime-specific dependencies still apply. Catalog templates, prompt packs and playbooks are separate reference documents.

SKILL.md — Standard Structure

---
name: example-skill
description: Explain what the skill does and when to use it.
---

# Example Skill

## Requirements
List required tools, credentials and supported platforms.

## Commands
| Command | Description |
|---------|-------------|
| skill-name do-thing | Does the thing |

## Example Output
```
Expected output here
```

## Notes
- Any caveats, auth requirements, or dependencies
- Cost/quota considerations

SKILL.md is the agent-readable interface contract. It tells the agent exactly what the skill does, how to call it, and what to expect back — no docs hunting required.

4. API Reference

Schema for /api/catalog-full.json — each SKU object.

Field	Type	Description
`slug`	`string`	URL-safe unique identifier. Use for /products/[slug] page.
`title`	`string`	Human-readable product name.
`short`	`string`	One-line summary (~100 chars). Use for catalog listings.
`long`	`string`	Full product description (Markdown). Use for product pages.
`price`	`number`	Price in USD. 0 = free, otherwise float (e.g. 1.00, 2.00).
`price_tier`	`string`	Tier label: "free" ($0), "micro" ($0.50), "standard" ($1), "pro" ($2+).
`category`	`string`	Product category: "skill" \| "prompt-pack" \| "playbook" \| "capability-extension" \| "bundle".
`runtime`	`string[]`	Runtime or document requirements, e.g. ["openclaw"], ["python3"], or ["markdown"].
`agent_use_case`	`string`	When to use this skill — the agent decision trigger.
`install_complexity`	`string`	"instant" (30 sec) or "5min" (requires API key setup).
`buy_url`	`string`	Empty for currently available free listings.
`tags`	`string[]`	Keyword tags for filtering and discovery.
`install`	`string`	Install command or instructions. Shown verbatim to agents.
`example_output`	`string`	What the agent sees after running the skill. Sets expectations.
`deliverables`	`string[]`	Bullet list of exactly what the buyer receives.

// Example SKU object
{
  "slug": "skill-github",
  "title": "GitHub Skill for OpenClaw",
  "short": "GitHub ops via gh CLI: issues, PRs, CI runs...",
  "long": "Full description...",
  "price": 0.5,
  "price_tier": "micro",
  "category": "skill",
  "runtime": ["openclaw"],
  "agent_use_case": "Use when an agent needs to interact with GitHub repos.",
  "install_complexity": "instant",
  "buy_url": "",
  "tags": ["github", "git", "pr", "ci", "devtools"],
  "install": "Bundled with OpenClaw; run openclaw skills check",
  "example_output": "PR #42: 'Fix login bug' — status: open, CI: passing",
  "deliverables": ["Install instructions...", "Copy-paste examples..."]
}

5. Case Studies

Real agent workflows using TutuoAI skills — installation and usage examples.

🚧

Coming Soon

Real agent workflows using tutuoai skills. In the meantime, each product page includes an example_output field showing what the agent receives after running the skill.