mcp-console
ctf-brain-user — participant client (MCP)
A stdio MCP server you connect to your CLI agent (Claude Code, Codex, Google Antigravity, or any MCP-capable agent) to work in your team's shared workspace.
Requirements
Node.js 22+
gitYour team's server URL and a personal API key (generate it on the Settings page).
Install
git clone https://github.com/sanjarbiy/ctf-brain-user
cd ctf-brain-user
npm install1. Add your keys (one-time)
On the Settings page, add at least 3 OpenRouter keys + 1 Jina key — or from the CLI:
node --import tsx client/src/cli.ts keys2. Generate your API key
On Settings → Connect API keys, generate a key (shown once — copy it). It carries your workspace, so BRAIN_CTF is not needed.
3. Connect the MCP to your agent
The MCP server is always the same command:
node --import tsx <ABSOLUTE-PATH>/ctf-brain-user/client/src/cli.ts mcpwith these env vars (use an absolute BRAIN_STATE_DIR — ~ is not expanded by every agent):
BRAIN_SERVER=https://your-team-server.example
BRAIN_TOKEN=mcp_your_generated_key
BRAIN_STATE_DIR=/home/you/.mcp-console # Windows: C:\Users\you\.mcp-console
BRAIN_VAULT_KEY=<64-hex> # openssl rand -hex 32Claude Code
claude mcp add mcp-console \
--env BRAIN_SERVER=https://your-team-server.example \
--env BRAIN_TOKEN=mcp_your_generated_key \
--env BRAIN_STATE_DIR=$HOME/.mcp-console \
--env BRAIN_VAULT_KEY=$(openssl rand -hex 32) \
-- node --import tsx /absolute/path/ctf-brain-user/client/src/cli.ts mcpRestart Claude Code.
Codex CLI
Option A — codex mcp add (everything after -- is the server command):
codex mcp add mcp-console \
--env BRAIN_SERVER=https://your-team-server.example \
--env BRAIN_TOKEN=mcp_your_generated_key \
--env BRAIN_STATE_DIR=/home/you/.mcp-console \
--env BRAIN_VAULT_KEY=<64-hex> \
-- node --import tsx /absolute/path/ctf-brain-user/client/src/cli.ts mcpOption B — edit ~/.codex/config.toml (Windows: C:\Users\you\.codex\config.toml):
[mcp_servers.mcp-console]
command = "node"
args = ["--import", "tsx", "/absolute/path/ctf-brain-user/client/src/cli.ts", "mcp"]
startup_timeout_sec = 60 # tsx compiles TS on first launch; raise from the default 10s
[mcp_servers.mcp-console.env]
BRAIN_SERVER = "https://your-team-server.example"
BRAIN_TOKEN = "mcp_your_generated_key"
BRAIN_STATE_DIR = "/home/you/.mcp-console"
BRAIN_VAULT_KEY = "<64-hex>"Set startup_timeout_sec = 60 even if you used Option A (edit the entry add created) — otherwise the first launch times out while tsx compiles. Restart Codex; verify with codex mcp list.
Google Antigravity ("agy")
Edit ~/.gemini/config/mcp_config.json (Windows: C:\Users\you\.gemini\config\mcp_config.json) — or in the IDE open … → MCP Servers → Manage MCP Servers → View raw config, or in the CLI type /mcp:
{
"mcpServers": {
"mcp-console": {
"command": "node",
"args": ["--import", "tsx", "/absolute/path/ctf-brain-user/client/src/cli.ts", "mcp"],
"env": {
"BRAIN_SERVER": "https://your-team-server.example",
"BRAIN_TOKEN": "mcp_your_generated_key",
"BRAIN_STATE_DIR": "/home/you/.mcp-console",
"BRAIN_VAULT_KEY": "<64-hex>"
}
}
}
}On Windows use forward slashes in the path (or escape backslashes). Save, then reload from the Manage MCP Servers panel (IDE) or /mcp (CLI).
hackerai
hackerai does not connect external MCP servers — it runs its own built-in tools in a sandbox. Two options to use this client from there:
Run it as a shell command in hackerai's terminal — install this repo on the sandbox host and call a CLI subcommand directly (see
cli.ts help), instead of the persistentmcpserver.Add it as a native tool by patching the hackerai source (advanced).
Safety
ctf_agent runs commands as local shell on your machine (guarded only against host-destructive commands like rm -rf /). Enable it only on a host where you accept local command execution. Make sure the client logs to stderr, not stdout — stdio MCP multiplexes JSON-RPC over stdout.
CLI
node --import tsx client/src/cli.ts helplogin / redeem (join with an invite code), keys (submit your OpenRouter/Jina keys), mcp (serve the MCP), status, doctor.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/sanjarbiy/ctf-brain-user'
If you have feedback or need assistance with the MCP directory API, please join our Discord server