Thronglets MCP
A P2P shared memory substrate for AI agents that creates a stigmergic knowledge network via libp2p. It allows coding agents to externalize logs, replays, and success/failure paths into sparse traces, which are then compressed into regressive experiences. The server injects sparse signals like 'avoid', 'do next', and 'maybe also' to guide agent decisions without explicit API calls, enabling collective intelligence and local learning across multiple sessions.
Package and access
Local package with source-attributed installation metadata. Koskamo does not execute packages during review.
What you can do with it
- Externalize agent logs and failure paths into regressive experiences
- Inject sparse decision signals like 'avoid' and 'do next' to guide agents
- Enable P2P shared memory for collective intelligence via libp2p
- Compress traces into space/field views for local agent learning
Source and maintenance
- Maintenance checked
Packages and installation
Source-provided package data. Koskamo has not executed or live-probed this package.
https://github.com/Shangri-la-0428/Thronglets/releases/download/v0.2.0/thronglets-mcp-linux-amd64 0.2.0
{
"version": "0.2.0",
"transport": "stdio",
"fileSha256": "d02883a6eecb861de8c1328ee9f264d4eb2d17635eb8db069bdf66ea7e9f33e6",
"identifier": "https://github.com/Shangri-la-0428/Thronglets/releases/download/v0.2.0/thronglets-mcp-linux-amd64",
"registryType": "mcpb",
"packageArguments": [
{
"type": "positional",
"value": "mcp"
}
]
}Tools
Reported by the attributed source; not independently observed by Koskamo.
Let your agent find its tools
Find servers like Thronglets MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.