High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.
MCP server and CLI for serial devices
Serial MCP Server gives agents and scripts a way to work with USB-serial and UART devices from either MCP tools or a direct CLI. It includes macro packs in JSON so you can validate, plan, simulate, and run timed serial workflows before touching hardware.
Builders who need their agent to inspect, script, or troubleshoot serial hardware.
You can automate repeatable serial workflows and run them through an agent without retyping each command and delay.
What it does
MCP serial tools
Exposes serial operations through MCP tools such as list ports, open, read, write, close, and set control lines.
Direct CLI
Provides command-line access for listing ports, probing devices, reading, writing, and changing RTS or DTR without an MCP client.
JSON macro automation
Defines repeatable serial procedures as JSON macros with `send`, `delay`, `expect`, and `assembly` steps.
Macro validation and planning
Lets you validate a macro pack, list its macros, expand a plan, and inspect the sequence before running it.
Dry runs and simulation
Supports dry-run and simulated reads so you can test macro logic without attached hardware.
Agent skill pack
Includes `skills/serial-debug/` for Claude Code and Codex users who want a reusable serial debugging skill.
How to get it
- 1Run
git clone https://github.com/adancurusul/serial-mcp-server.git cd serial-mcp-server cargo build --release
- 2The binary is built at
target/release/serial-mcp-server
- 3To install the CLI onto your PATH from a checkout
cargo install --path . --locked
- 4Use --duration-ms when an AI or script needs one bounded collection window. Capture mode…
serial-mcp-server read --port <port> --baud 115200 \ --duration-ms 5000 \ --start-trigger first-byte \ --initial-timeout-ms 30000 \ --idle-timeout-ms 1500 \ --max-bytes 8192 \ --json
- 5write --read accepts the same capture options after the write
serial-mcp-server write --port <port> --baud 115200 --data RUN --read \ --duration-ms 5000 --initial-timeout-ms 30000 --json
- 6Configuration commands
serial-mcp-server generate-config serial-mcp-server validate-config --config serial-mcp.toml serial-mcp-server show-config --config serial-mcp.toml
README
Serial MCP Server
serial-mcp-server provides serial port access for AI workflows in two forms:
- MCP stdio server for clients that support MCP tools.
- Scriptable CLI for direct use, CI, and agent skills without MCP setup.
Current release target: 0.3.0.
Language versions: English | Chinese
0.3.0 Update Brief
- Added JSON Macro DSL automation for repeatable serial procedures.
- Added CLI macro commands:
macro validate,macro list,macro plan, andmacro run. - Added MCP macro tools for runtime in-memory pack load/list/unload, plan, and run.
- Added explicit no-hardware simulation for macro validation, planning, and executor smoke tests.
- Kept Quick out of the public API. Quick-style use cases should be represented as macros.
Macro Automation
Use macros when a serial workflow is more than a single read or write. Many devices require a timed sequence: send a command, wait a few milliseconds, read until a prompt or acknowledgement appears, then send the next command. A macro pack records that procedure as JSON so a human, CLI script, or AI agent can validate it, inspect the plan, simulate it without hardware, and run it against a real port when a device is attached.
Typical macro use cases:
- Boot or provisioning flows that need ordered commands and delays.
- Protocol handshakes that must wait for
OK,READY,PONG, prompts, or other expected responses. - Regression smoke tests where the same serial procedure should run repeatedly.
- AI-assisted debugging where the agent should review the full send/delay/expect plan before touching hardware.
The v0.3 DSL is intentionally small:
send: write UTF-8, hex, or base64 bytes.delay: wait for a fixed number of milliseconds.expect: read until the response contains or equals expected bytes.assembly: compose named macros into a longer workflow.
AI agents can discover macro support from this README, from the bundled skills/serial-debug skill, by running serial-mcp-server macro --help, or through MCP tool discovery when the server is configured. Agents that do not use MCP can still use the CLI plus the skill docs.
Requirements
- Rust 1.74 or newer.
- A serial device or USB-to-serial adapter when running hardware operations.
- Device drivers and OS permissions for the selected serial port.
Install From Source
git clone https://github.com/adancurusul/serial-mcp-server.git
cd serial-mcp-server
cargo build --release
The binary is built at:
target/release/serial-mcp-server
To install the CLI onto your PATH from a checkout:
cargo install --path . --locked
CLI Usage
Use the CLI when you want direct serial operations without an MCP client.
serial-mcp-server --help
serial-mcp-server list-ports --json
serial-mcp-server probe --port <port> --baud 115200 --json
serial-mcp-server write --port <port> --baud 115200 --data H --read --timeout-ms 1000 --json
serial-mcp-server read --port <port> --baud 115200 --timeout-ms 1000 --json
serial-mcp-server read --port <port> --baud 115200 --duration-ms 5000 --initial-timeout-ms 30000 --idle-timeout-ms 1500 --json
serial-mcp-server set-control-lines --port <port> --rts high --dtr low --json
Capture Window Reads
timeout-ms keeps its original one-read meaning: wait up to that many
milliseconds for one read operation, then return as soon as bytes are available.
Use --duration-ms when an AI or script needs one bounded collection window.
Capture mode keeps reading internally and returns one combined response with
completion_reason, waited_ms, elapsed_ms, and per-chunk metadata.
serial-mcp-server read --port <port> --baud 115200 \
--duration-ms 5000 \
--start-trigger first-byte \
--initial-timeout-ms 30000 \
--idle-timeout-ms 1500 \
--max-bytes 8192 \
--json
Capture options:
--duration-ms: collection window length after capture starts.--start-trigger first-byte: wait for the first byte before starting the duration clock. This is the default for capture mode.--start-trigger immediate: start the duration clock when the command starts.--initial-timeout-ms: maximum wait for the first byte in first-byte mode. If omitted, the command uses--timeout-ms.--idle-timeout-ms: stop after this many quiet milliseconds once capture has started.--max-bytes: hard cap on the combined response bytes.
write --read accepts the same capture options after the write:
serial-mcp-server write --port <port> --baud 115200 --data RUN --read \
--duration-ms 5000 --initial-timeout-ms 30000 --json
Macro automation commands:
serial-mcp-server macro validate --file examples/macros/ping.json --json
serial-mcp-server macro list --file examples/macros/ping.json --json
serial-mcp-server macro plan --file examples/macros/ping.json --macro ping --json
serial-mcp-server macro run --file examples/macros/ping.json --macro ping --dry-run --json
serial-mcp-server macro run --file examples/macros/ping.json --macro ping --simulate-read PONG --json
serial-mcp-server macro run --file examples/macros/ping.json --macro ping --port <port> --baud 115200 --json
Macro packs are JSON files with schema_version set to 0.3. v0.3 supports send, delay, and expect steps inside macros, plus assemblies that call macros by name. expect supports contains and equals.
The macro DSL is intentionally restricted. It does not run shell commands, JavaScript, Python, file operations, loops, variables, if/else branches, Quick commands, or RTS/DTR macro steps.
Configuration commands:
serial-mcp-server generate-config
serial-mcp-server validate-config --config serial-mcp.toml
serial-mcp-server show-config --config serial-mcp.toml
CLI output rules:
- stdout is reserved for command data and JSON.
- stderr is reserved for diagnostics.
--jsonoutput should be parseable by tools such asjq.- Nonzero exit codes indicate command failure.
Supported CLI data formats are utf8, hex, and base64. Use hex or base64 for binary payloads.
MCP Usage
Use MCP when your client supports MCP tools and you want a long-running stdio server.
Recommended server command:
serial-mcp-server serve
No-subcommand startup is retained as a compatibility path for existing MCP setups, but new configurations should use serve.
Claude Desktop example for macOS/Linux:
{
"mcpServers": {
"serial": {
"command": "/path/to/serial-mcp-server/target/release/serial-mcp-server",
"args": ["serve"],
"env": {
"RUST_LOG": "info"
}
}
}
}
Windows example:
{
"mcpServers": {
"serial": {
"command": "C:\\path\\to\\serial-mcp-server\\target\\release\\serial-mcp-server.exe",
"args": ["serve"],
"env": {
"RUST_LOG": "info"
}
}
}
}
MCP Tools
| Tool | Purpose |
|---|---|
list_ports | Discover available serial ports. |
open | Open a serial connection. |
write | Write UTF-8, hex, or base64 data to an open connection. |
read | Read data from an open connection with timeout handling or a bounded capture window. |
close | Close an open connection. |
set_control_lines | Set RTS and/or DTR on an open connection. |
macro_load | Validate and load an inline macro pack or pack file path into the server's in-memory registry. |
macro_list | List loaded macro packs, macros, and assemblies. |
macro_unload | Remove a loaded macro pack from the in-memory registry. |
macro_plan | Expand a loaded, inline, or file-backed macro or assembly without opening hardware. |
macro_run | Run a loaded macro or assembly against an existing connection or explicit simulation input. |
macro_run_inline | Validate, plan, and run an inline macro pack without storing it in the registry. |
The MCP macro registry is runtime-only. Restarting the server clears loaded packs, and the server does not write a persistent macro library.
MCP read accepts these optional capture fields in addition to
connection_id, timeout_ms, max_bytes, and encoding:
{
"connection_id": "...",
"duration_ms": 5000,
"start_trigger": "first_byte",
"initial_timeout_ms": 30000,
"idle_timeout_ms": 1500,
"max_bytes": 8192,
"encoding": "utf8"
}
When duration_ms is absent, MCP read keeps the existing single-read
behavior. When duration_ms is present, the tool returns structured JSON text
with completion_reason, waited_ms, elapsed_ms, and chunks.
Agent Skill
The repository includes a Claude Code and Codex compatible skill at:
skills/serial-debug/
The skill is CLI-first and documents MCP as an optional configured path. It is intended for agents that need to list ports, run serial smoke tests, run macro automation, control RTS/DTR, or troubleshoot UART/USB-serial devices.
For local development, copy the skill folder into the agent skill roots:
mkdir -p ~/.codex/skills ~/.claude/skills ~/.agents/skills
cp -R skills/serial-debug ~/.codex/skills/
cp -R skills/serial-debug ~/.claude/skills/
cp -R skills/serial-debug ~/.agents/skills/
Tested explicit triggers:
Codex: Use $serial-debug
Claude Code: /serial-debug
Claude Code --bare mode did not resolve /serial-debug in local testing; use normal Claude Code print or interactive mode for skill-trigger smoke tests.
Hardware Safety
Serial commands can affect real hardware.
- Confirm the selected port from
serial-mcp-server list-ports --json. - Confirm voltage levels before connecting an adapter to a target board.
- Confirm baud rate, data bits, parity, stop bits, and flow control before writing.
- Treat RTS and DTR carefully. Many boards wire those lines to reset or boot mode.
- Validation claims should be based on command output from the connected device.
STM32 Demo
The STM32 demo is under:
examples/STM32_demo/
It provides firmware for an interactive serial command interface. See examples/STM32_demo/README.md for wiring, firmware commands, MCP usage, and CLI smoke commands.
Quality Gates
Release work uses the checked-in Cargo.lock and these gates:
cargo fmt --all -- --check
cargo clippy --locked --all-targets --all-features -- -D warnings
cargo test --locked --all-targets --all-features
cargo doc --locked --all-features --no-deps
CLI smoke:
cargo run --locked -- --help
cargo run --locked -- list-ports --json
cargo run --locked -- write --help
cargo run --locked -- set-control-lines --help
cargo run --locked -- macro validate --file examples/macros/ping.json --json
cargo run --locked -- macro run --file examples/macros/ping.json --macro ping --simulate-read PONG --json
License
This project is licensed under the MIT License. See LICENSE.
Files in the repo
- doc
- examples
- skills
- src
- tests
- .gitignore
- Cargo.lock
- Cargo.toml
- CHANGELOG.md
- CLAUDE.md
- LICENSE
- README_ZH.md
- README.md
- SERIAL_MCP_SERVER_THEORY_OF_OPERATION.md
Discussion (0)
Ask about usage, or say what you built with itSign in to join the discussion.
No comments yet. Be the first to say what this is good for.
More connectors

Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama…) with Codex CLI, App, SDK, and Claude Code
Git-native persistent memory for AI coding agents. Implements Google OKF v0.2 with sub-300µs in-memory BM25 search, embedded MCP server, and progressive disclosure. Slashes token bloat by 80% with zero external databases or dependencies. Built in pure Go.
Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked context reductions on reviews and large-repo workflows.
Stop your AI from making things up — it proposes, deterministic tools decide, every claim checked against ground truth with evidence. Grounded facts and context survive resets. Reverse engineering is the proving ground. MCP server + CLI.
20 MB lightweight cross-platform database client for 90+ databases, including MySQL, PostgreSQL, SQLite, Redis, MongoDB, DuckDB, SQL Server, and Dameng. Built-in AI, MCP Server, CLI, desktop and Docker. | 轻量级跨平台数据库管理工具,支持 MySQL、PostgreSQL、SQLite、Redis、MongoDB、达梦等 90+ 数据库,提供桌面端、Docker、CLI、内置 AI 助手和 MCP Server。