cortex-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| CORTEX_MAX_SESSIONS | No | Maximum number of active sessions | 100 |
| CORTEX_SESSION_TTL_MS | No | Session time-to-live in milliseconds (default 30 minutes) | 1800000 |
| CORTEX_MAX_TOTAL_TOKENS | No | Maximum total tokens across all sessions | 2000000 |
| CORTEX_REDACT_TRACE_CONTENT | No | Whether to redact trace content | false |
| CORTEX_MAX_ACTIVE_REASONING_TASKS | No | Maximum number of active reasoning tasks | 32 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
| prompts | {
"listChanged": true
} |
| resources | {
"listChanged": true
} |
| completions | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| reasoning_thinkA | Structured multi-step reasoning tool. Decomposes analysis into sequential thought steps stored in a persistent session trace. USAGE PATTERN:
IMPORTANT: Pass the returned sessionId on every continuation call. The thought parameter stores YOUR reasoning verbatim — write thorough analysis in each step. Use stepSummary for a 1-sentence conclusion per step — these accumulate in the summary field for navigation. Levels: basic (1–3 steps, 2K budget), normal (4–8 steps, 8K budget), high (10–15 steps, 32K budget), expert (20–25 steps, 128K budget). Alternatives: runMode="run_to_completion" (batch, basic level only), or observation/hypothesis/evaluation fields (structured). Errors: E_SESSION_NOT_FOUND (expired — start new), E_INVALID_THOUGHT_COUNT (check level ranges). |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| reasoning.basic | Basic-depth reasoning (1-3 thoughts). |
| reasoning.normal | Normal-depth reasoning (4-8 thoughts). |
| reasoning.high | High-depth reasoning (10-15 thoughts). |
| reasoning.expert | Expert-depth reasoning (20-25 thoughts). |
| reasoning.retry | Retry a failed reasoning task with modified parameters. |
| get-help | Return server usage instructions. |
| reasoning.continue | Continue an existing session. Optional follow-up query. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| server-instructions | Usage instructions for the MCP server. |
| tool-catalog | Tool reference: models, params, outputs, data flow. |
| workflows | Recommended workflows and tool sequences. |
| server-config | Runtime limits and level configurations for the reasoning server. |
| reasoning.sessions | List of active reasoning sessions with summaries. Updated in real-time as sessions progress. |
| tool-info-reasoning_think | Contract details for reasoning_think. |
TDQS
Scored across 1 tool
With only one tool, there is no risk of confusing it with other tools. The tool's purpose is clearly defined as structured multi-step reasoning.
The single tool name 'reasoning_think' follows a clear verb_noun pattern (verb 'think' with noun 'reasoning' as modifier), and consistency is not an issue with only one tool.
One tool is minimal but appropriate given the focused domain of reasoning. The tool is complex with multiple levels and modes, so it earns its place, but the count feels slightly thin for a general MCP server.
The tool covers the full lifecycle of multi-step reasoning: start, continue, complete, with support for different levels and modes. No obvious gaps for a reasoning-only server.