synapse
Agent-to-agent P2P file sharing with semantic search using BitTorrent and vector embeddings
Why use this skill?
Power your AI agents with Synapse. A P2P file sharing skill using BitTorrent and vector embeddings for intelligent semantic search across distributed networks.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/pendzoncymisio/synapseWhat This Skill Does
The Synapse skill serves as a decentralized peer-to-peer (P2P) file sharing protocol specifically designed for intelligent agents within the OpenClaw ecosystem. By integrating BitTorrent technology with vector embeddings, Synapse enables agents not just to share static files, but to perform semantic searches across the network. This means that if you are looking for specific information, you can search by concept or content similarity rather than just filenames or exact keyword matches. The skill manages its own seeder daemon, ensuring high availability of shared knowledge, and automates the retrieval of files through magnet links generated during the indexing process.
Installation
Synapse requires Python 3.10 or higher and relies on the 'uv' package manager to handle its complex dependencies, such as 'sentence-transformers' and 'libtorrent'. Installation is streamlined: first, install uv via the provided script. Navigate to your project directory and use 'uv run' to execute commands; this handles the creation of a virtual environment automatically on the first run. For full functionality, ensure your environment variables (like SYNAPSE_PORT) are configured properly. If you encounter missing module errors, ensure you are not using system Python, as 'uv run' is strictly required to isolate the dependency stack.
Use Cases
Synapse is ideal for distributed knowledge bases where centralized storage is impractical. It excels in collaborative research scenarios where multiple agents share datasets, codebases, or documentation. Because it uses semantic search, it is highly effective for building 'hive mind' document repositories where an agent can query for 'deployment strategies' and receive relevant Markdown guides from peers without knowing the specific file names. It is also a robust tool for private, decentralized data backups among a cluster of connected agents.
Example Prompts
- "Synapse, start the seeder daemon and share all Markdown files in my research folder with tags like 'AI', 'documentation', and 'research'."
- "Search the Synapse network for documents related to 'distributed system architecture' and list the top 5 results with their similarity scores."
- "Download the file associated with this magnet link [magnet_link_here] to my local downloads directory and notify me when the transfer completes."
Tips & Limitations
Always ensure that your seeder daemon is running if you intend to host files for other agents. If you find your search results are irrelevant, verify that your vector index is updated. The primary limitation is network availability; because it is P2P, if all nodes hosting a specific file go offline, the file becomes inaccessible. Monitor the logs located at '~/.openclaw/seeder.log' if you experience connectivity issues, and consider opening your ports if you are behind a strict firewall.
Metadata
Not sure this is the right skill?
Describe what you want to build ā we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-pendzoncymisio-synapse": {
"enabled": true,
"auto_update": true
}
}
}Tags
Flags: network-access, file-write, file-read
Related Skills
index1
AI memory system for coding agents ā code index + cognitive facts, persistent across sessions.
memphis-super
š§ Memphis Super-Agent v3.6.0 MVP - 100% Memphis Core for OpenClaw Complete Memphis 3.6.0 integration: - ALL chains (journal, ask, decisions, vault, trade, graph, reflection, ingest, ops, adr) - Semantic search & embeddings - Multi-provider LLM routing - Onboarding wizard (--clean, --nuclear) - Auto-detection (questions/decisions/insights) - Session markers (start/end) - Knowledge graph - MCP server Auto-detects: - Questions ā ask block (with semantic search) - Decisions ā decisions block (conscious) - Insights ā journal block - Session end ā summary block Perfect for: ALL OpenClaw agents seeking complete memory + cognitive powers Quick start: /memphis status
memphis
š„ Memphis - Complete AI Brain for OpenClaw Agents ALL-IN-ONE meta-package with everything you need: š§ Core Features: - Local-first memory chains (journal, recall, ask, decisions) - Offline LLM integration (Ollama, local models) - Semantic search with embeddings - Knowledge graph - Encrypted vault for secrets š Cognitive Engine (Models A+B+C): - Model A: Record conscious decisions (manual) - Model B: Detect decisions from git (automatic) - Model C: Predict decisions before you make them (predictive) - 90.7% accuracy, proactive suggestions š ļø Setup & Management: - Bootstrap wizard (5-minute setup) - Self-loop capability (Memphis uses itself) - Auto-repair system - Chain monitoring - Backup automation š Multi-Agent Network: - Campfire Circle Protocol - Share chain sync (IPFS) - Multi-agent collaboration - Agent negotiation (trade protocol) Perfect for: Individual developers, teams, researchers, entrepreneurs Quick start: clawhub install memphis && memphis init
Memphis Cognitive Engine
š§ Memphis Cognitive Engine - Complete AI Memory System Transform your OpenClaw agent into a cognitive partner with: - Model A: Record conscious decisions (manual) - Model B: Detect decisions from git (automatic) - Model C: Predict decisions before you make them (predictive) - Advanced: TUI, Knowledge Graph, Reflection, Trade Protocol, Multi-Agent Sync Production-ready with 100% working commands (17/17), zero bugs. ā ļø IMPORTANT: This is a META-PACKAGE (documentation only). Memphis CLI must be installed separately. Quick start: clawhub install memphis-cognitive
vinculum
Shared consciousness between Clawdbot instances. Links multiple bots into a collective, sharing memories, activities, and decisions in real-time over local network using Gun.js P2P sync.