# cognition-process

`station__cognition__cognition-process` · external (needs EXECUTION_BACKEND_URL configured) · domain `cognition` · pv-relevance `non-pv`

(for composability with cognition_perplexity), and attention profile.

> **Note:** This tool routes through an external execution backend. If `EXECUTION_BACKEND_URL` is unset on the server, calls return JSON-RPC error `-32603 "Tool execution backend not configured"`. Tools with `backend: native` execute in-process and are always callable.

## Agent metadata

- `idempotent`: unknown
- `read_only`: unknown
- `expected_latency_ms`: unknown (not yet contract-tested)
- `cost_tokens_estimate`: unknown

## Input schema

- `ffn_inner_dim` *integer*
- `max_new_tokens` *integer* (required)
- `max_seq_len` *integer* (required)
- `min_confidence` *number*
- `model_dim` *integer* (required)
- `num_heads` *integer* (required)
- `num_layers` *integer* (required)
- `prompt` *array* (required) — Token IDs representing the input prompt.
- `repetition_penalty` *number* (required)
- `seed` *integer*
- `stop_token` *integer*
- `temperature` *number* (required)
- `top_k` *integer* (required)
- `top_p` *number* (required)
- `vocab_size` *integer* (required)

## Example call

```json
POST /api/mcp
Content-Type: application/json

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "station__cognition__cognition-process",
    "arguments": {
      "max_new_tokens": 0,
      "max_seq_len": 0,
      "model_dim": 0,
      "num_heads": 0,
      "num_layers": 0,
      "prompt": [],
      "repetition_penalty": 0,
      "temperature": 0,
      "top_k": 0,
      "top_p": 0,
      "vocab_size": 0
    }
  }
}
```

## Related

- [/tools](/tools) — all 3062 tools
- [/tools/cognition__cognition-process](/tools/cognition__cognition-process) — HTML page
- [/tools/cognition__cognition-process/json](/tools/cognition__cognition-process/json) — JSON form (agent-friendly)
- [/api/mcp](/api/mcp) — endpoint
- [/AGENTS.md](/AGENTS.md) — agent guide