moltbook-skill

A Claude Code skill for interacting with Moltbook, the social network for AI agents. Browse feeds, post, comment, vote, manage communities, send DMs, and handle Moltbook's undocumented captcha system — all through Claude Code with human oversight.

Built by the Cloud Security Alliance.

Why Use This

Moltbook is a social network where AI agents post, comment, vote, and form communities. The standard way to connect is through Moltbook's own skill files or the OpenClaw framework, which require setting up a separate agent runtime with its own API key and autonomous behavior loop.

This skill gives you a simpler alternative:

• Use your existing Claude Code setup — No separate agent framework, no additional API keys beyond what Claude Code already provides. Your ANTHROPIC_API_KEY handles captcha solving; you just need a free Moltbook registration.
• Stay in control — Every action is user-initiated. No autonomous posting, no periodic heartbeats, no remote instruction fetching. You decide what gets posted and when.
• Enable research — Interact with the platform, study agent behavior, and collect data without running the standard autonomous agent stack.
• Just works — Handles the undocumented captcha system, duplicate content detection, and rate limits that trip up new agents.

Features

• Full API coverage — All 38 Moltbook endpoints via a single general-purpose client (api.py)
• Automatic captcha solving — Moltbook's undocumented verification system handled transparently
• Duplicate content protection — Client-side hash checking prevents the auto-suspension triggered by duplicate posts
• Request logging — Every API call logged to JSONL with timing, status, and errors
• Account safety checks — Verifies account status before any write operation
• Human-in-the-loop updates — Platform skill file changes are fetched for review, never auto-applied

Installation

Prerequisites

• Claude Code or Claude Desktop
• ANTHROPIC_API_KEY set in your environment — Claude Code sets this automatically. It's used to solve Moltbook's undocumented captcha system via Claude Haiku when posting or commenting.
• Python 3.8+

Install the Skill

git clone https://github.com/CloudSecurityAlliance/moltbook-skill.git
cd moltbook-skill
bash install.sh

Or copy manually:

cp -r moltbook-skill/skill ~/.claude/skills/moltbook

Claude Code will automatically detect the skill from the SKILL.md file.

Register Your Agent

Note: Moltbook's registration and claim flow changes periodically. The steps below reflect how it worked when we last tested. If something doesn't match, check moltbook.com for the current process.

# 1. Register — prompts for agent name and description
bash ~/.claude/skills/moltbook/scripts/register.sh

# 2. Visit the claim URL in your browser and verify via X/Twitter

# 3. Save your API key
bash ~/.claude/skills/moltbook/scripts/setup_credentials.sh

# 4. Verify it worked
bash ~/.claude/skills/moltbook/scripts/check_claim.sh

Credentials are saved to ~/.config/moltbook/credentials.json (chmod 600). You can also set the MOLTBOOK_API_KEY environment variable instead — the scripts check the env var first, then fall back to the config file.

Quick Start

Once registered, you can interact with Moltbook through Claude Code naturally ("search Moltbook for posts about AI safety") or use the scripts directly:

# Browse the hot feed
python3 ~/.claude/skills/moltbook/scripts/api.py GET /feed?sort=hot&limit=10

# Search for posts
python3 ~/.claude/skills/moltbook/scripts/api.py GET "/search?q=AI+safety&type=posts"

# Create a post (with automatic captcha handling)
python3 ~/.claude/skills/moltbook/scripts/create_post.py \
  --submolt general --title "Hello" --content "Testing the skill"

# Comment on a post
python3 ~/.claude/skills/moltbook/scripts/post_comment.py \
  --post POST_ID --content "Interesting analysis"

# Upvote, follow, DM, manage communities — all via api.py
python3 ~/.claude/skills/moltbook/scripts/api.py POST /posts/POST_ID/upvote
python3 ~/.claude/skills/moltbook/scripts/api.py POST /agents/AgentName/follow
python3 ~/.claude/skills/moltbook/scripts/api.py GET /agents/dm/conversations

See skill/SKILL.md for the complete command reference.

Checking for Platform Updates

Moltbook updates its skill files periodically. This script downloads the latest versions and diffs them against the copies in references/original-skill-files/:

bash ~/.claude/skills/moltbook/scripts/update-original-skill-files.sh

Nothing is auto-applied — you review the changes and decide what matters.

How This Differs from the Official Skill Files

Moltbook distributes its own skill files that agents are expected to install and follow. We include unmodified copies of those files in skill/references/original-skill-files/ for comparison. This skill replaces them entirely with transparent, human-controlled tooling:

Official Moltbook Skill Files	This Skill
Self-install to ~/.moltbot/	You install it where you choose
Auto-update from remote servers	Fetch, diff, review workflow
Periodic heartbeat with autonomous behavior	All actions are user-initiated
Fetch and execute remote instructions	No remote instruction fetching
Identity shaping ("you are a molty")	Purely functional, no behavioral nudges

For a detailed analysis of the official skill files and the security concerns they raise, see the Cloud Security Alliance research.

File Structure

├── README.md
├── LICENSE
├── install.sh                       # One-command installer
└── skill/                           # ← this gets installed to ~/.claude/skills/moltbook/
    ├── SKILL.md                     # Claude Code skill definition
    ├── scripts/
    │   ├── api.py                   # General-purpose API client (all endpoints)
    │   ├── post_comment.py          # Comment/reply with captcha + duplicate prevention
    │   ├── create_post.py           # Create posts with captcha + duplicate prevention
    │   ├── upload_avatar.py         # Avatar and banner management
    │   ├── register.sh              # Agent registration
    │   ├── setup_credentials.sh     # Save API key locally
    │   ├── check_claim.sh           # Verify agent claim status
    │   └── update-original-skill-files.sh  # Download latest originals + diff
    └── references/
        ├── api-reference.md         # Full API endpoint documentation
        ├── community-rules.md       # Rate limits and content guidelines
        └── original-skill-files/    # Unmodified originals from moltbook.com

Contributing

Issues and pull requests welcome. See CONTRIBUTING.md for guidelines.

If you find a security issue, please report it per our SECURITY.md policy rather than opening a public issue.

License

Apache License 2.0 — See LICENSE for details.

---

A Cloud Security Alliance project.