GPT Commander
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| CODEX_PATH | No | Path to codex CLI (optional) | |
| CLAUDE_PATH | No | Path to Claude CLI (optional) | |
| GEMINI_PATH | No | Path to Gemini CLI (optional) | |
| GPT_DC_HOST | No | Host for SSE MCP (default 127.0.0.1) | 127.0.0.1 |
| GPT_DC_PORT | No | Port for SSE MCP (default 3333) | 3333 |
| GPT_DC_CONFIG | No | Path to custom config file (optional) | |
| GEMINI_WORKDIR | No | Work directory for Gemini CLI | |
| GPT_DC_TLS_KEY | No | Path to TLS key for HTTPS | |
| GPT_DC_TLS_CERT | No | Path to TLS cert for HTTPS | |
| GPT_DC_TAILNET_HOST | No | Allowed host list for SSE | |
| GPT_DC_ABSOLUTE_ENDPOINT | No | Absolute base URL for SSE rewrite |
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 |
|---|---|
| commander_onboardA | Start mandatory onboarding and return the one sheet + obstacle course |
| commander_finishB | Complete onboarding after required steps |
| get_metricsC | Get telemetry metrics and recent events |
| memory_noteC | Store a note in the memory MCP |
| list_directoryB | List directory entries |
| read_fileD | Read a text file with limits |
| write_fileC | Write a text file (line-limited) |
| get_file_infoD | Stat a file |
| start_processC | Start a long-running process |
| interact_with_processC | Send stdin to a process |
| read_process_outputC | Read buffered process output |
| list_sessionsB | List active processes |
| kill_processC | Kill a process |
| start_searchC | Search files with ripgrep |
| edit_blockB | Replace the Nth occurrence of text in a file |
| ide_list_workspacesA | List connected IDE workspaces |
| ide_get_diagnosticsB | Get IDE diagnostics (optionally filtered by file or severity) |
| ide_symbol_searchC | Search workspace symbols using the IDE index |
| ide_find_referencesC | Find references using the IDE index |
| ide_read_bufferB | Read an IDE buffer (includes unsaved changes when available) |
| ide_apply_workspace_editC | Apply a VS Code WorkspaceEdit (atomic) with a preview summary |
| project_infoC | Detect project tooling, package managers, scripts, and CI |
| run_testsC | Run project tests with a detected command |
| run_unit_testsC | Run unit tests with a detected command |
| run_integration_testsC | Run integration tests with a detected command |
| lintC | Run lint with a detected command |
| formatC | Run formatting with a detected command |
| typecheckC | Run typecheck with a detected command |
| buildC | Run build with a detected command |
| dev_server_startC | Start a detected dev server and return a process id |
| codex_runC | Run the Codex CLI with safe arguments |
| claude_runC | Run the Claude CLI with safe arguments |
| gemini_runC | Run the Gemini CLI with safe arguments |
| dev_server_stopA | Stop a dev server by process id |
| git_statusC | Get git status (short) |
| git_diffC | Get git diff (optionally staged) |
| git_commit_prepareC | Stage changes (optional) and return a commit prep summary |
| search_codeC | Search code using ripgrep with safe defaults |
| find_symbolC | Find symbols using IDE index (LSP) |
| open_file_at_lineC | Read a file snippet around a line number |
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 40 tools
Several tools have overlapping purposes: start_search and search_code both use ripgrep, and ide_symbol_search and find_symbol both use the IDE/LSP symbol index. This creates ambiguity in tool selection, especially for agents trying to choose the right search or symbol tool.
Most tools follow a verb_noun snake_case pattern (e.g., read_file, kill_process, list_sessions, dev_server_start). However, memory_note is a noun phrase rather than a verb, and a few tools are bare verbs (lint, format, build), creating minor inconsistency.
At 40 tools, the surface is very large. While the server aims to cover many domains, there are redundant tools (e.g., start_search/search_code, ide_symbol_search/find_symbol) that inflate the count and could be consolidated, making the tool set feel heavy and less focused.
Core workflows such as file operations, process management, IDE integration, and project commands are well covered. However, there are notable gaps: no file delete/rename, no memory retrieval (only memory_note), and git tools stop short of an actual commit, leaving some workflows incomplete.