Skip to main content
Glama
chacharin

mcp-thai-proofread

by chacharin

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
RUN_GIDNoGID to run as; should match the owner of the agent workspace.
RUN_UIDNoUID to run as; should match the owner of the agent workspace.
WORKERSNoNumber of concurrent LLM requests.
LLM_MODELYesModel name to use for Thai proofreading.
CHUNK_CHARSNoMaximum number of characters per document chunk before splitting.
LLM_API_KEYYesAPI key for the LLM endpoint. Must be changed from the sample value; required if the endpoint uses authentication.
LLM_BASE_URLYesBase URL of the OpenAI-compatible LLM endpoint.
NOTES_SUBDIRNoSubdirectory name for proofread notes. Set empty to disable notes; backup files cannot be disabled.
PUBLISH_ADDRNoAddress the server publishes on. Default is 0.0.0.0; use 127.0.0.1 to restrict to localhost.
BACKUP_SUBDIRNoSubdirectory name for backup files. Default is backups.
CONTEXT_LINESNoNumber of lines before and after a chunk provided as read-only context to the LLM.
PATH_PREFIXESNoPath prefixes to strip from agent-provided file paths, for example /opt/data/workspace/.
DOCKER_NETWORKNoDocker network name to use. Defaults to agent_network.
MCP_AUTH_TOKENNoToken for authenticating MCP requests. Recommended; if unset, anyone on the network can modify workspace files.
MIN_SIMILARITYNoMinimum similarity threshold for accepting an edited line.
FILE_EXTENSIONSNoAllowed file extensions for proofreading.
PROTECTED_REGEXNoRegex for protected codes that must not change during proofreading.
LLM_EXTRA_PARAMSNoJSON string of provider-specific LLM options, such as top_p or OpenRouter provider selection.
OUTPUT_PLACEMENTNoPlacement for backups and notes: file (same directory as edited file, default) or top (first-level folder under workspace).
SYSTEM_PROMPT_FILENoPath to a custom system prompt file, for example /app/prompts/system_prompt.md.
WORKSPACE_HOST_DIRNoHost directory to mount as the workspace. Defaults to ./workspace.
LLM_TIMEOUT_SECONDSNoLLM request timeout in seconds. MCP client timeout should not be set lower than this value.

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

Server capabilities have not been inspected yet.

Tools

Functions exposed to the LLM to take actions

NameDescription

No tools

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources