input.send_actions
PREFERRED multi-step tool: run 1–10 predictable UI actions in one call (input.click_target, input.click_xy, input.type_text, input.send_keys, input.drag, input.scroll). Side effects: all actions execute on the remote desktop; fails fast before sending if any action is invalid. observe_after defau...
This record as markdown: /tools/com-glasswarp-mcp-server/input.send-actions.md
What input.send_actions does on Glasswarp
AI agents invoke input.send_actions to trigger actions in Glasswarp. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call: builds kicked off, notifications sent, workflows started.
| Parameter | Type | Required | Description |
|---|---|---|---|
actions | array | Yes | Ordered actions (max 10); input.click_target needs target_id from latest screen.observe |
session_id | string | Yes | Active session id |
observe_after | boolean | — | Default true. When true, return a verification observe in the same result |
observe_image | boolean | — | Default false. When true with observe_after, include verification JPEG unless changed=false |
Parameters from the server's own tool schema.
Why input.send_actions is rated High
This tool triggers external operations (UI automation on a remote Windows PC) whose effects depend entirely on the arguments provided. An AI agent could use this to launch applications, navigate to malicious sites, enter credentials into wrong systems, or trigger unintended desktop actions. While individual actions are reversible, the combination and sequence create unpredictable side effects on a real system.
From the tool's definition Tool executes multiple UI actions on remote desktop: 'run 1–10 predictable UI actions in one call', 'all actions execute on the remote desktop'. Supports click_target, click_xy, type_text, send_keys, drag, scroll — all direct desktop manipulation operations.
Risk signalsHigh parameter count (16 properties)
Attacks that exploit this kind of access
The rule that runs input.send_actions safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Glasswarp, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For input.send_actions, this is the rule to start with:
input.send_actions stays usable, but rate-capped: a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Glasswarp, apply this rule, and every input.send_actions call is checked against it from then on.
Questions about input.send_actions
PREFERRED multi-step tool: run 1–10 predictable UI actions in one call (input.click_target, input.click_xy, input.type_text, input.send_keys, input.drag, input.scroll). Side effects: all actions execute on the remote desktop; fails fast before sending if any action is invalid. observe_after defaults true (verification observe: text+targets; set observe_image=true for JPEG). Do not batch across unpredictable waits (page loads, installers, modals) — single-step those. Prefer this over chaining solo click/type/keys tools. It is categorised as a Execute tool in the Glasswarp MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
input.send_actions accepts 4 parameters: actions, session_id, observe_after, observe_image. Required: actions, session_id. The full parameter table on this page comes from the server's own tool schema.
Register the Glasswarp MCP server in PolicyLayer and add a rule for input.send_actions: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Glasswarp. Nothing to install.
input.send_actions is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the input.send_actions rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for input.send_actions. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
input.send_actions is provided by the Glasswarp MCP server (https://mcp.glasswarp.com/mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Glasswarp, and thousands of servers like it.
Across the catalogue