sidecar-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| SIDECAR_MODEL | No | Model to use for the worker. Default depends on backend: ollama -> llama3.1:8b, openai -> gpt-4o-mini, anthropic -> claude-3-5-haiku-latest. | |
| SIDECAR_BACKEND | No | Backend to use for the worker model. One of: ollama, openai, anthropic, fake. Default: ollama. | ollama |
| SIDECAR_LOG_LEVEL | No | Log level. One of: error, info, debug. Default: info. | info |
| SIDECAR_OLLAMA_URL | No | URL of the Ollama server. Default: http://127.0.0.1:11434. | http://127.0.0.1:11434 |
| SIDECAR_OPENAI_KEY | No | API key for OpenAI. Required when SIDECAR_BACKEND=openai. | |
| SIDECAR_OPENAI_URL | No | URL of an OpenAI-compatible endpoint. Default: https://api.openai.com. | https://api.openai.com |
| SIDECAR_ALLOW_ROOTS | No | Comma-separated absolute paths to restrict file access. Default: cwd (current working directory). | cwd |
| SIDECAR_ANTHROPIC_KEY | No | API key for Anthropic. Required when SIDECAR_BACKEND=anthropic. Also accepts Anthropic-compatible providers. | |
| SIDECAR_ANTHROPIC_URL | No | URL of an Anthropic-compatible endpoint. Default: https://api.anthropic.com. | https://api.anthropic.com |
| SIDECAR_FILE_MAX_BYTES | No | Maximum file size in bytes to read. Larger files are skipped and reported. Default: 524288 (512 KB). | 524288 |
| SIDECAR_REQUEST_TIMEOUT_MS | No | Timeout for worker requests in milliseconds. Default: 120000. | 120000 |
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
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| bulk_readA | Read multiple files and ask the worker LLM to summarize them in the context of |
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
With only one tool, there is no possibility of confusion or overlap with other tools. The tool's purpose is clearly described.
A single tool name cannot demonstrate a pattern, but 'bulk_read' is a clear verb_noun construction. The naming is sensible, though consistency cannot be assessed across a set.
A single tool is extremely thin for a server, even if it serves a specific niche. The server's purpose appears to be reading and summarizing files, which likely requires additional supporting tools for full utility.
The tool covers bulk reading and summarization, but there are no complementary tools for file listing, individual reads, or other common operations. The surface is severely limited for a file-access server.