Velvet Unicorn Skills Library

Public repository of skills for Moltbot (formerly Clawdbot) — including the Velvet Unicorn skill for AI-powered token, wallet, and trading analysis via the VelvetDAO Unicorn /ask endpoint.

Structure

Each top-level directory is a provider. Each subdirectory within a provider is an installable skill containing a SKILL.md and other skill-related files.

moltbot-skills/
└── velvet/
    └── velvet-unicorn/
        ├── SKILL.md
        └── scripts/
            ├── ask_unicorn.py
            └── ask_unicorn.sh

Install Instructions

Give Moltbot the URL to this repo and it will let you choose which skill to install.

https://github.com/<YOUR_ORG>/<YOUR_REPO>

Available Skills

Provider	Skill	Description
velvet	velvet-unicorn	Calls VelvetDAO Unicorn AI (POST /ask) to answer token analysis, wallet analysis, and trade analysis questions. Returns { meta_category, vu_category, answer, txdata }.

Skill Overview: velvet-unicorn

The velvet-unicorn skill is a thin wrapper around:

• Endpoint: POST https://gentweet.velvetdao.xyz/ask

• Purpose: Provide natural-language answers for:

  • Token information & analysis
  • Wallet analysis
  • Trading / “should I buy/sell” analysis (may return tx intent data)

Request Format

The helper scripts send JSON shaped like:

{
  "question": "string",
  "context": "string",
  "userid": "string | null",
  "userid_sol": "string | null",
  "previous_questions": ["string", "..."]
}

Response Format

The endpoint is expected to return:

{
  "meta_category": 0,
  "vu_category": 1,
  "answer": "string",
  "txdata": { "..." : "..." }
}

• answer is the text response you show the user.
• txdata may be null or {}; if present, treat as proposed transaction data (not executed).

Usage

Bash wrapper

bash velvet/velvet-unicorn/scripts/ask_unicorn.sh \
  "Analyze $TOKEN and summarize risk + momentum" \
  "User is interested in short-term swing trades; medium risk tolerance." \
  "0xYourEvmWallet" \
  "" \
  '["compare $TOKEN to $OTHER","what is the downside risk?"]'

Python script

python3 velvet/velvet-unicorn/scripts/ask_unicorn.py \
  --question "Analyze $TOKEN and summarize risk + momentum" \
  --context "User is interested in short-term swing; medium risk tolerance." \
  --userid "0xYourEvmWallet" \
  --userid-sol "" \
  --previous-questions '["compare $TOKEN to $OTHER","what is the downside risk?"]'

Environment Variables

• VELVET_UNICORN_ENDPOINT (optional) — override default endpoint (https://gentweet.velvetdao.xyz/ask)
• VELVET_UNICORN_API_KEY (optional) — if your endpoint requires auth, sent as Authorization: Bearer <key>

Contributing

Community contributions are welcome. Here’s how to add your own skill:

Adding a New Skill

1. Fork this repository and create a new branch for your skill.
2. Create a provider directory (if it doesn’t exist):

mkdir your-provider-name/

3. Add the required files:

   • SKILL.md — the main skill definition (required)
   • references/ — supporting docs (optional)
   • scripts/ — helper scripts (optional)

4. Follow the structure:

your-provider-name/
└── your-skill-name/
    ├── SKILL.md
    ├── references/
    │   └── your-docs.md
    └── scripts/
        └── your-script.sh

5. Submit a Pull Request with a clear description of your skill.

Guidelines

• Keep skill definitions clear and well-documented
• Include usage examples in SKILL.md
• Test your skill before submitting
• Use descriptive commit messages