ESP Input Bridge
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| ESP_HID_TOKEN | No | Shared HID token for hardware mode, matching the token provisioned on the ESP32 board. Required when using hardware mode; not needed for demo mode. | |
| ESP_INPUT_CONFIG | No | Path to the TOML configuration file. If unset, the server uses ~/.config/esp-input-bridge/config.toml when present; with no configuration it runs in demo mode. |
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 | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| observeB | Read target state and a fresh observation ID; optionally include a cropped image. |
| actB | Execute up to 8 hardware input actions on a fresh observation. Return next state/image. |
| stopA | Stop this session and release inputs; restart the server to permit inputs again. |
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 3 tools
Each tool has a clearly distinct role: observe reads state, act performs input actions, and stop ends the session. There is no meaningful overlap in intent or selection risk.
All three names are short lowercase imperative verbs with a consistent style. The convention is predictable across the entire tool set.
Three tools is a tight, well-scoped surface for an input bridge: observe, act, and stop correspond directly to the core interaction loop. Each tool earns its place, with no redundant or filler operations.
The observe-act-stop loop covers basic operation, but stopping is a dead end unless the server is restarted externally. There is no explicit start/reset tool and no way to discover supported action types, which are notable gaps for agent use.