Skip to main content
Glama
Node Mantis

Node Mantis

Official

Run a website task

run_task
Destructive

Carry out a website task from a plain-English goal in a real browser: navigate, click, type, fill forms, and read the outcome, returning structured data when a schema is supplied.

Instructions

Carry out a task on a website from a plain-English goal: navigate, click, type, fill in forms and read the outcome, in a real browser. Returns the outcome and, when result_schema is given, the data in that shape. If the task is still going when the call returns, the reply has status "running" and a task_id: call get_task with it. Each call starts a fresh browser session with no saved logins.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "goal": "<goal>"
    },
    "name": "run_task"
  }
}

Input Schema

NameRequiredDescriptionDefault
goalYesWhat to accomplish, with every detail needed to finish it, e.g. "Find the cheapest direct flight from Boston to Denver on March 3 and return the airline, times and price".
effortNoHow much capability to spend. Default medium; raise it for hard pages, lower it for simple ones.
secretsNoNames of private values the task may type (passwords, one-time codes). Each value is read from NODEMANTIS_SECRET_<NAME> in this server's environment; refer to one in the goal as <secret:NAME>.
max_stepsNoStop after this many steps. Default 30.
start_urlNoPage to open first (http or https). Omit to let the task find its own way.
use_proxyNoRoute the browser through the proxy configured in this server's environment. Default false.
max_cost_usdNoStop once the task has spent this many US dollars.
show_browserNoShow the browser window while the task runs. Default false unless the server was started with --headed.
result_schemaNoJSON Schema describing the shape of the data to return, for example {"type":"object","properties":{"price":{"type":"number"}},"required":["price"]}. Omit it for a free-form answer.
timeout_secondsNoStop after this long. Default 300.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.1.9

TDQS

A4.1/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The annotations mark this as not read-only, destructive, non-idempotent and open-world, and the description adds behaviour they cannot express: each browser session is fresh with no saved logins, the reply can come back early with status 'running' and a task_id, and results are shaped by result_schema. That is solid added context, although it does not say what kinds of state or side effects in the outside world to expect from an action that is flagged destructive.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

It is front-loaded: purpose first, then the return shape, then the running/get_task handoff, then the session caveat. Every sentence does useful work, with no filler or repetition, though the four-sentence block could be broken into clearer statements.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 10-parameter, open-world, stateful tool with no output schema, the definition covers how to call it, how the result is shaped, the asynchronous result flow, and the session and credential assumptions. The destructive and side-effect profile is only partly spelled out, and the default behaviour for findings such as timing and cost is left to the schema, so it is near-complete but not exhaustive.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already explains all ten parameters, including effort levels, secrets from NODEMANTIS_SECRET_<NAME>, max_steps, start_url and the result_schema contract. The description only restates the result_schema effect and adds the credential note about <secret:NAME>, which is useful but thin over what the schema already covers, so the baseline of 3 holds.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description gives a specific verb and resource ('Carry out a task on a website from a plain-English goal') and enumerates the concrete actions performed (navigate, click, type, fill in forms, read the outcome) in a real browser. The agent can cleanly distinguish it from get_task, which is named as the follow-up.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It clearly says when to use it (a plain-English goal to execute on a website) and routes the agent to get_task when the reply has status 'running' plus a task_id. The remaining siblings, cancel_task and retrieve_data, are not mentioned, so guidance on when not to use it or how to cancel or retrieve data is missing.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server

Other Tools