
proxyman-mcp-setup
Connect an AI coding agent to Proxyman MCP. Use when Proxyman is installed but the agent cannot see tools, or when configuring Codex, Claude, Cursor, VS Code, or Copilot and troubleshooting bridge, handshake, or tool-discovery errors.
- Installs
- 440
- Last 24h
- —
- Trending rank
- —
- Security
- Not audited
Install
Works with every agent the skills CLI supports.
npx skills add https://github.com/proxymanapp/proxyman-skill.md --skill proxyman-mcp-setup- 01
Install
Run the command in your project. The CLI asks which agents to install for.
- 02
Describe the task
Ask your agent for something this skill covers. It loads proxyman-mcp-setup automatically when relevant.
- 03
Review the result
The agent follows the SKILL.md steps, scripts and references bundled with the skill.
SKILL.md
The instructions your agent reads when the skill is triggered.
Proxyman MCP Setup
Configure an agent to talk to Proxyman MCP through Proxyman's bundled stdio bridge.
Operating Rules
- This skill is shell-first. MCP tools may not be connected yet.
- Do not use API-key or direct HTTP MCP configuration. Proxyman MCP uses a local stdio bridge executable.
- Proxyman must be installed, running, and have Settings > MCP > MCP Server enabled before verification can pass.
- Preserve existing MCP server entries in agent config files. Add or update only the
proxymanentry. - Prefer the exact command shown in Proxyman Settings > MCP when the app exposes one.
Mental Model
Proxyman MCP has two local pieces:
- The AI agent launches Proxyman's bundled
mcp-serverexecutable over stdio. - The bridge reads
mcp-handshake.jsonfrom Proxyman's app data folder. - The bridge forwards tool calls to the running app at
http://127.0.0.1:<ephemeral-port>/mcpwith a bearer token from the handshake file. - The bridge belongs to one client session and exits when the client's stdin reaches EOF. The MCP client should launch it again for the next session.
Do not hardcode the HTTP port or token. The app regenerates them.
Common handshake locations:
| Platform/build | Handshake folder |
|---|---|
| macOS native app | Proxyman Application Support bundle folder, including regular and Setapp bundle IDs |
| Windows Electron app | %APPDATA%\Proxyman |
| Linux Electron app | ${XDG_CONFIG_HOME:-$HOME/.config}/Proxyman |
Do not edit the handshake file. If it is missing or stale, restart Proxyman and re-enable Settings > MCP.
Step 1: Verify Proxyman Is Installed
macOS
if [ -d "/Applications/Proxyman.app" ]; then
echo "INSTALLED: /Applications/Proxyman.app"
elif mdfind 'kMDItemCFBundleIdentifier == "com.proxyman.NSProxy"' | grep -q "Proxyman.app"; then
mdfind 'kMDItemCFBundleIdentifier == "com.proxyman.NSProxy"'
else
echo "NOT_INSTALLED"
fi
Windows (PowerShell)
$installed = Get-ItemProperty `
"HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\Uninstall\*",
"HKCU:\SOFTWARE\Microsoft\Windows\CurrentVersion\Uninstall\*",
"HKLM:\SOFTWARE\WOW6432Node\Microsoft\Windows\CurrentVersion\Uninstall\*" `
-ErrorAction SilentlyContinue |
Where-Object { $_.DisplayName -like "*Proxyman*" }
if ($installed) { "INSTALLED" } else { "NOT_INSTALLED" }
Linux
if command -v proxyman >/dev/null 2>&1 || ls "$HOME"/Downloads/Proxyman*.AppImage "$HOME"/Downloads/proxyman*.AppImage >/dev/null 2>&1; then
echo "INSTALLED"
else
echo "NOT_INSTALLED"
fi
If Proxyman is not installed, stop and use proxyman-download-setup.
Step 2: Launch Proxyman And Enable MCP
macOS
open -a "Proxyman"
sleep 10
Windows (PowerShell)
$candidates = @(
"$env:LOCALAPPDATA\Programs\Proxyman\Proxyman.exe",
"C:\Program Files\Proxyman\Proxyman.exe",
"C:\Program Files (x86)\Proxyman\Proxyman.exe"
)
$proxymanExe = $candidates | Where-Object { Test-Path $_ } | Select-Object -First 1
if ($proxymanExe) { Start-Process $proxymanExe; Start-Sleep 10 }
Linux
if command -v proxyman >/dev/null 2>&1; then
nohup proxyman >/dev/null 2>&1 &
else
nohup "$HOME/Downloads/Proxyman.AppImage" >/dev/null 2>&1 &
fi
sleep 10
Ask the user to open Proxyman Settings > MCP and enable "MCP Server". If the MCP toggle is locked, the user must authorize the plan/license required by their Proxyman build.
If a bundled proxyman-cli is already available, it can persist the same state even while Proxyman is closed. Read proxyman-cli mcp --help from that installed build, then use its mcp on action only when the user asked to enable MCP. Launch Proxyman afterward before expecting operational MCP calls. Use proxyman-cli skill for executable discovery and exact syntax; do not guess flags from this setup guide.
Keep "Redact Sensitive Data Before Sending to AI" enabled unless the user explicitly wants raw headers, cookies, query strings, or bodies sent to the agent.
On Linux AppImage builds, launching Proxyman lets the app copy the packaged MCP bridge into a stable config path and mark it executable. If the bridge path below is missing, enable MCP in Settings, restart Proxyman, and check again.
Step 3: Resolve The Bridge Executable
macOS
Regular build:
BRIDGE_PATH="/Applications/Proxyman.app/Contents/MacOS/mcp-server"
test -x "$BRIDGE_PATH" && echo "$BRIDGE_PATH"
If the regular path is missing, search installed Proxyman apps:
mdfind 'kMDItemCFBundleIdentifier == "com.proxyman.NSProxy" || kMDItemCFBundleIdentifier == "com.proxyman.NSProxy-setapp"' |
while read -r app; do
candidate="$app/Contents/MacOS/mcp-server"
[ -x "$candidate" ] && echo "$candidate"
done
Windows (PowerShell)
The Windows Electron app copies mcp-server.exe beside Proxyman.exe. If no bridge is found, use the exact command from Proxyman Settings > MCP.
$exeCandidates = @(
"$env:LOCALAPPDATA\Programs\Proxyman\Proxyman.exe",
"C:\Program Files\Proxyman\Proxyman.exe",
"C:\Program Files (x86)\Proxyman\Proxyman.exe"
)
$proxymanExe = $exeCandidates | Where-Object { Test-Path $_ } | Select-Object -First 1
$bridgeCandidates = @()
if ($proxymanExe) {
$bridgeCandidates += Join-Path (Split-Path $proxymanExe -Parent) "mcp-server.exe"
}
$bridgeCandidates += @(
"$env:LOCALAPPDATA\Programs\Proxyman\mcp-server.exe",
"C:\Program Files\Proxyman\mcp-server.exe",
"C:\Program Files (x86)\Proxyman\mcp-server.exe"
)
$bridge = $bridgeCandidates | Where-Object { Test-Path $_ } | Select-Object -First 1
if ($bridge) { $bridge } else { "NOT_FOUND: copy the path from Proxyman Settings > MCP" }
Linux
The packaged Linux AppImage copies mcp-server into Proxyman's config folder after launch. Prefer that stable copied path.
CONFIG_HOME="${XDG_CONFIG_HOME:-$HOME/.config}"
BRIDGE_PATH="$CONFIG_HOME/Proxyman/bin/mcp-server"
if [ -x "$BRIDGE_PATH" ]; then
echo "$BRIDGE_PATH"
else
echo "NOT_FOUND: launch Proxyman, enable Settings > MCP, restart Proxyman, or copy the path from Settings > MCP"
fi
Set BRIDGE_PATH to the chosen executable path. It must be the stdio bridge, not the Proxyman app binary.
Step 4: Detect The Agent Config
Use environment variables first, then parent process, then filesystem markers.
[ -n "$OPENAI_CODEX" ] && echo "codex"
[ -n "$CLAUDE_CODE_ENTRYPOINT" ] && echo "claude-code"
[ -n "$CURSOR_TRACE_ID" ] || [ "$TERM_PROGRAM" = "cursor" ] && echo "cursor"
[ -n "$VSCODE_PID" ] || [ "$TERM_PROGRAM" = "vscode" ] && echo "vscode"
[ -n "$GITHUB_COPILOT_CLI" ] && echo "copilot-cli"
Fallback markers:
test -d "$HOME/.codex" && echo "codex"
test -d "$HOME/.claude" && echo "claude-code"
test -d "$HOME/.cursor" && echo "cursor"
test -d ".vscode" && echo "vscode"
test -f "$HOME/.copilot/mcp-config.json" && echo "copilot-cli"
test -f "$HOME/Library/Application Support/Claude/claude_desktop_config.json" && echo "claude-desktop"
If multiple agents are detected, ask the user which agent they want to configure.
Step 5: Add The MCP Server
Codex CLI
Preferred command:
codex mcp add proxyman -- "$BRIDGE_PATH"
Equivalent TOML:
[mcp_servers.proxyman]
enabled = true
command = "BRIDGE_PATH"
args = []
Config file: ~/.codex/config.toml.
Claude Code
claude mcp add proxyman --transport stdio -- "$BRIDGE_PATH"
Config file: ~/.claude.json.
Claude Desktop
Config file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
Server entry:
{
"mcpServers": {
"proxyman": {
"command": "BRIDGE_PATH",
"args": [],
"env": {}
}
}
}
Cursor
Config file: ~/.cursor/mcp.json.
{
"mcpServers": {
"proxyman": {
"command": "BRIDGE_PATH",
"args": [],
"env": {}
}
}
}
VS Code / GitHub Copilot
Config file:
- macOS:
~/Library/Application Support/Code/User/mcp.jsonor workspace.vscode/mcp.json - Linux:
~/.config/Code/User/mcp.jsonor workspace.vscode/mcp.json - Windows:
%APPDATA%\Code\User\mcp.jsonor workspace.vscode\mcp.json
{
"servers": {
"proxyman": {
"command": "BRIDGE_PATH",
"args": [],
"env": {}
}
}
}
GitHub Copilot CLI
Config file: ~/.copilot/mcp-config.json.
{
"mcpServers": {
"proxyman": {
"command": "BRIDGE_PATH",
"args": [],
"env": {},
"tools": ["*"]
}
}
}
When editing JSON or TOML config manually, parse and merge with a real parser when possible. Never replace unrelated mcpServers or servers entries.
Step 6: Verify
Restart or reload the agent after changing config.
Verification sequence:
- Confirm Proxyman is running.
- Confirm Settings > MCP shows the server running.
- Ask the agent to list MCP tools/resources/prompts if it supports discovery.
- Call
get_version. - Call
get_proxy_status.
Successful setup means the agent can see Proxyman tools and get_version returns a Proxyman app/bridge response.
Tool availability can vary by platform and build. Treat the live tool list as authoritative. In particular, macOS system proxy control, Apple Terminal/iTerm/Ghostty injection, .app Electron injection, and bridge app-control tools are macOS-oriented; do not offer them on Windows or Linux unless that server exposes an applicable schema.
Troubleshooting
| Error | Meaning | Action |
|---|---|---|
Proxyman is not running or MCP server not started | Bridge cannot find mcp-handshake.json. | Launch Proxyman and enable Settings > MCP. |
Invalid handshake file | Token or port is stale. | Restart Proxyman, then reload the agent. |
Cannot connect to Proxyman | App is closed or MCP server stopped. | Open Proxyman and confirm MCP status. |
| Linux bridge path is missing | The AppImage has not prepared mcp-server in the config folder yet. | Launch Proxyman, enable Settings > MCP, restart Proxyman, then check ${XDG_CONFIG_HOME:-$HOME/.config}/Proxyman/bin/mcp-server. |
| Agent has no Proxyman tools | Config path is wrong or agent was not reloaded. | Re-check BRIDGE_PATH, config file, and restart the agent. |
| User sees no traffic after setup | MCP is connected but target capture is not configured. | Use proxyman-https-capture, then inspect with proxyman-traffic-debugging. |
Next Step
Once get_version and get_proxy_status work:
- use
proxyman-traffic-debuggingfor the complete MCP operation surface; - use
proxyman-https-capturefor device, emulator, browser, runtime, and framework routing/trust setup; - use
proxyman-debugging-toolsfor current official feature documentation; - use
proxyman-clifor version-matched shell automation.
FAQ
What does the proxyman-mcp-setup skill do?
Connect an AI coding agent to Proxyman MCP. Use when Proxyman is installed but the agent cannot see tools, or when configuring Codex, Claude, Cursor, VS Code, or Copilot and troubleshooting bridge, handshake, or tool-discovery errors.
How do I install proxyman-mcp-setup?
Run npx skills add https://github.com/proxymanapp/proxyman-skill.md --skill proxyman-mcp-setup in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/proxyman-mcp-setup/ (personal) or .claude/skills/proxyman-mcp-setup/ (project).
Which AI agents can use proxyman-mcp-setup?
Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.
Is proxyman-mcp-setup safe to install?
No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.
Who maintains proxyman-mcp-setup?
It is published by proxymanapp in proxymanapp/proxyman-skill.md. It has 440 recorded installs.
More from proxymanapp/proxyman-skill.md
proxyman-traffic-debugging
proxymanapp/proxyman-skill.md
Use Proxyman MCP to inspect and debug HTTP, HTTPS, and WebSocket traffic and operate the complete Proxyman MCP surface. Use for flows, API diagnosis, filtering, export, replay, Compose, debugging rules, WebSockets, recording, proxy and certificate control, code generation, or guided setup.
proxyman-download-setup
proxymanapp/proxyman-skill.md
Open the skill to read its SKILL.md instructions.
proxyman-debugging-tools
proxymanapp/proxyman-skill.md
Find and explain current Proxyman debugging tools using official documentation, with GUI, MCP, and CLI workflows where supported. Use for Breakpoint, Map Local or Remote, Scripting, Compose, Repeat, Protobuf, TLS Key Logging, Reverse Proxy, Network Conditions, WebSocket, filters, exports, certificates, and other Proxyman feature guides.
proxyman-https-capture
proxymanapp/proxyman-skill.md
Configure and troubleshoot Proxyman HTTPS capture across desktop systems, Apple and Android devices or simulators, browsers, runtimes, containers, frameworks, and HTTP clients, including Atlantis instrumentation for controlled apps.
proxyman-cli
proxymanapp/proxyman-skill.md
Discover and safely use the installed proxyman-cli for proxy control, MCP state, licenses, configuration, logs, certificates, rules, and debugging-tool automation. Use for Proxyman shell commands, scripts, CLI help, or operations that MCP does not execute.
proxyman-app-settings
proxymanapp/proxyman-skill.md
Open the skill to read its SKILL.md instructions.
Popular ai & agents skills
google-agents-cli-publish
google/agents-cli
This skill should be used when the user wants to "publish an agent", "publish my ADK agent", "register an agent with Gemini", "publish to Gemini" (formerly Gemini Enterprise), or needs guidance on the agents-cli publish gemini command. Also use when the user wants to "manage agents in Agent Registry", "list/update/delete registered agents", or "register an MCP server". Covers ADK vs A2A registration modes, programmatic and interactive usage, flag reference, auto-detection from deployment metadata, Agent Registry fleet management, and troubleshooting. Part of the agents-cli skills suite. Do NOT use for deployment (use google-agents-cli-deploy).
google-agents-cli-eval
google/agents-cli
This skill should be used when the user wants to "run an evaluation", "evaluate my agent", "evaluate my ADK agent", "write an eval dataset", "analyze eval failures", "compare eval results", "optimize agent", or needs guidance on the Gemini platform eval methodology and the Quality Flywheel. Covers eval metrics, dataset schema, LLM-as-judge scoring, and common failure causes. Applies to any agents-cli project, whatever framework the agent is written in. Do NOT use for agent API code patterns (ADK: use google-agents-cli-adk-code), deployment (use google-agents-cli-deploy), or project scaffolding (use google-agents-cli-scaffold).
skill-creator
anthropics/skills
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
codex-pet
prime-skills/runcomfy-agent-skills
Codex Pet generator on RunComfy. Build a Codex-compatible Codex Pet spritesheet.webp + pet.json from a single reference image, drop it into ${CODEXHOME:-$HOME/.codex}/pets/ / and Codex picks it up as a custom Codex Pet next to the 8 built-ins. This skill produces the exact Codex Pet atlas Codex expects (1536x1872 PNG/WebP, 8 cols x 9 rows, 192x208 cells, 9 animation states — idle, running-right, running-left, waving, jumping, failed, waiting, running, review). Calls OpenAI GPT Image 2 edit ONCE via the local RunComfy CLI as runcomfy run openai/gpt-image-2/edit to produce a canonical Codex Pet pose, then assembles all 9 animation rows programmatically with ImageMagick micro-transforms — no Codex Pro, no $imagegen, no OPENAIAPIKEY required, only RUNCOMFYTOKEN. Triggers on "codex pet", "create codex pet", "make codex pet", "hatch codex pet", "/hatch image", "desktop pet codex", "codex pets", "spritesheet.webp", or any explicit ask to build a custom pet for OpenAI Codex.
full-output-enforcement
leonxlnx/taste-skill
Overrides default LLM truncation behavior. Enforces complete code generation, bans placeholder patterns, and handles token-limit splits cleanly. Apply to any task requiring exhaustive, unabridged output.
executing-plans
obra/superpowers
Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available