CLI Agent MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| CAM_GUI | No | Enable GUI dashboard | true |
| CAM_DEBUG | No | Include debug info in MCP responses | false |
| CAM_TOOLS | No | Comma-separated list of allowed tools (empty = all) | |
| CAM_GUI_KEEP | No | Keep GUI on exit | false |
| CAM_LOG_DEBUG | No | Write debug logs to temp file | false |
| CAM_GUI_DETAIL | No | GUI detail mode | false |
| CAM_SIGINT_MODE | No | SIGINT handling (cancel/exit/cancel_then_exit) | cancel |
| CAM_SIGINT_DOUBLE_TAP_WINDOW | No | Double-tap exit window (seconds) | 1.0 |
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
| Name | Description |
|---|---|
| codexA | Run OpenAI Codex CLI agent (deep analysis / critical review). NO SHARED MEMORY:
CROSS-AGENT HANDOFF:
CAPABILITIES:
BEST PRACTICES:
Supports: image attachments. |
| codex_parallelA | Run multiple codex tasks in parallel. All tasks share workspace/permission/save_file. Results are appended to save_file with XML wrappers (). Max 100 tasks. Model can be array: single element shared by all, or one per task. |
| geminiA | Run Google Gemini CLI agent (UI design / comprehensive analysis). NO SHARED MEMORY:
CROSS-AGENT HANDOFF:
CAPABILITIES:
BEST PRACTICES:
|
| gemini_parallelA | Run multiple gemini tasks in parallel. All tasks share workspace/permission/save_file. Results are appended to save_file with XML wrappers (). Max 100 tasks. Model can be array: single element shared by all, or one per task. |
| claudeA | Run Anthropic Claude CLI agent (code implementation). NO SHARED MEMORY:
CROSS-AGENT HANDOFF:
CAPABILITIES:
BEST PRACTICES:
Supports: system_prompt, append_system_prompt, agent parameter. |
| claude_parallelA | Run multiple claude tasks in parallel. All tasks share workspace/permission/save_file. Results are appended to save_file with XML wrappers (). Max 100 tasks. Model can be array: single element shared by all, or one per task. |
| opencodeA | Run OpenCode CLI agent (full-stack development). NO SHARED MEMORY:
CROSS-AGENT HANDOFF:
CAPABILITIES:
BEST PRACTICES:
Supports: file attachments, multiple agents (build, plan, etc.). |
| opencode_parallelA | Run multiple opencode tasks in parallel. All tasks share workspace/permission/save_file. Results are appended to save_file with XML wrappers (). Max 100 tasks. Model can be array: single element shared by all, or one per task. |
| bananaA | Generate images using Nano Banana Pro (Gemini 3 Pro Image). CAPABILITIES:
RESPONSE FORMAT:
BEST PRACTICES:
Supports: reference images with roles (edit_base, style_ref, etc.). |
| imageA | Generate images via OpenRouter-compatible or OpenAI-compatible endpoints. CAPABILITIES:
RESPONSE FORMAT:
BEST PRACTICES:
Supports: reference images for editing. |
| get_gui_urlA | Get the GUI dashboard URL. Returns the HTTP URL where the live event viewer is accessible. |
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 11 tools
Most tools have distinct purposes targeting different AI agents or functions (e.g., banana for image generation, claude for code implementation, get_gui_url for dashboard access). However, there is some overlap between 'banana' and 'image' as both handle text-to-image generation and editing, which could cause confusion despite different backend providers.
The naming is mixed: some tools use simple names (banana, image, claude, codex, gemini, opencode), while others append '_parallel' for parallel execution variants (claude_parallel, codex_parallel, gemini_parallel, opencode_parallel). This pattern is somewhat consistent for parallel tools, but the base names lack a uniform verb_noun structure, making it less predictable overall.
With 11 tools, the count is reasonable for a CLI agent server that orchestrates multiple AI agents and utilities. It covers core agents, their parallel versions, and auxiliary functions like image generation and GUI access, though it might be slightly heavy if some tools are rarely used.
The tool set provides good coverage for running various AI agents (claude, codex, gemini, opencode) with parallel execution options, plus image generation and GUI access. A minor gap is the lack of tools for managing agent sessions or configurations, but core workflows are well-supported.