CotForce-MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MODE | No | auto, sampling, or direct. auto uses direct HTTP when API_KEY is set and client lacks sampling support. | auto |
| MODEL | No | Model name hint (e.g. claude-3-5-sonnet, gpt-4o). If empty, no hint sent – MCP host decides. | |
| API_KEY | No | LLM API key for direct HTTP mode. Optional for local endpoints, required for remote providers. | |
| TIMEOUT | No | Sampling timeout in ms (60s for sampling, 120s for direct HTTP). | 60000 / 120000 |
| BASE_TEMP | No | Initial sampling temperature. | 0.1 |
| CACHE_TTL | No | Result cache TTL in ms (default 1 hour). Set to 0 to disable. | 3600000 |
| LOG_LEVEL | No | One of DEBUG, INFO, WARN, ERROR. | INFO |
| COT_PARSERS | No | Comma-separated parser names to use (e.g., direct-json,fenced-block). Skips others. | |
| MAX_RETRIES | No | Number of retry attempts before returning raw output. | 2 |
| API_BASE_URL | No | Base URL for direct HTTP mode. Default is https://api.openai.com. | https://api.openai.com |
| TEMP_INCREMENT | No | Temperature added per retry attempt. | 0.2 |
| FALLBACK_MODELS | No | Comma-separated list of fallback models (e.g. gpt-4o,claude-3-5-sonnet). Cycled on failure. | |
| CACHE_MAX_ENTRIES | No | Maximum cached results before evicting oldest. | 100 |
| REASONING_OVERHEAD | No | Fixed token overhead added to the budget formula. | 800 |
| TRUNCATION_THRESHOLD | No | Ratio of output/budget that triggers truncation detection. | 0.95 |
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 | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| solve_problemA | Solves a problem using strict Agentic Chain-of-Thought with adaptive parsing, retry logic, token budgeting (via tiktoken), and configurable model. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 1 tool
Only one tool exists, so there is no possibility of confusion with other tools.
With a single tool, naming is trivially consistent; the name 'solve_problem' is clear and descriptive.
The server has only one tool, which is on the thin side. For a server promising broad problem-solving capabilities, additional tools might be expected, but the singular focus could be justified.
The single tool covers only one operation. There are no tools for configuration, error handling, or feedback, leaving significant gaps in the expected surface for a comprehensive problem-solving server.