# Jaysonaionline Cognitive Substrate MCP server

Source: https://glimind.com/server/mcp-registry%2Fio.github.jaysonaionline%2Fcognitive-substrate
Registry id: `mcp-registry/io.github.jaysonaionline/cognitive-substrate`
Status: unknown

## Install

```json
{
  "mcpServers": {
    "jaysonaionline-cognitive-substrate-mcp-s": {
      "command": "uvx",
      "args": [
        "mcp-cognitive-substrate"
      ]
    }
  }
}
```

Claude Code:

```sh
claude mcp add jaysonaionline-cognitive-substrate-mcp-s -- uvx mcp-cognitive-substrate
```

## Tools (10)

- `a2a_call`: Call a peer agent with a message.
- `a2a_discover`: Discover capabilities of a peer agent.
- `a2a_list`: List available peer agents for A2A communication.
- `a2a_orchestrate`: Orchestrate a task across peer agents based on capability.
- `get_cognitive_tree_state`: Return active paths and pruned branches for a session's ToT evolutionary memory.
- `prune_failed_approach`: Prune a failed approach node from the ToT memory, turning it into a lesson.
- `reason`: Apply a reasoning paradigm (deductive, inductive, abductive, analogical, causal, syllogistic, falsification) to a problem.
- `run_paradigm`: Run a single reasoning paradigm from the 28-layer substrate.
- `store_5key_telemetry`: Store compact 5-key telemetry (foundations, metacognition, defensive, resource, utility) for a session.
- `substrate_run`: Validate a prompt through the 28-layer cognitive substrate and return the verdict.

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.
