Skip to main content
Glama

Ask Cursor (new chat)

cursor_ask

Run a new Cursor coding chat with an optional model and project workspace to answer questions or perform agentic coding tasks. Continue the same task with follow-up calls.

Instructions

Best for agentic coding on a Cursor plan, with a wide model menu (GPT/Claude/Grok/Composer via model) (cursor-agent -p), in a NEW chat. Pass workspace=; without it the sub-agent sees no repo. Follow up with cursor_continue. Uses your Cursor login or CURSOR_API_KEY. Its sandbox is agent-enforced (read-only = ask mode), not an OS boundary. Autonomous agent: trusted prompts only.

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": {
      "prompt": "<prompt>"
    },
    "name": "cursor_ask"
  }
}

Input Schema

NameRequiredDescriptionDefault
modelNoOptional model override (`--model`, e.g. "auto", "gpt-5.2", "claude-opus-4-8-high", "composer-2.5"); validated against `cursor-agent models` and rejected on a typo. cursor bakes the effort and speed axes into the id (…-low/-high/-xhigh/-max, each with a -fast twin) and also accepts a bracket form on the family base, e.g. "claude-opus-4-8[context=1m,effort=high]". Omit to use your Cursor default.
watchNoIf true, open a live "watch" view streaming cursor's steps from its `--output-format stream-json` event stream. Same final text is returned. Best-effort. Default false.
promptYesQuestion or instruction for Cursor.
sandboxNoPermission policy (maps to cursor's mode/force flags): "read-only" (default — `--mode ask`: agent-enforced, no file/shell edits; NOT an OS sandbox, so unlike codex it is not a hard boundary), "workspace-write" (may edit files, rooted at the workspace), or "danger-full-access" (OS sandbox off — avoid).read-only
timeout_sNoMax seconds to wait for cursor to complete. Default 180.
workspaceNoWorking root for the chat (`--workspace`). Defaults to the server cwd.

Output Schema

NameRequiredDescriptionDefault
resultYes

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changedv0.22.1
    • changedInput schema / properties / model / description
      Previous value: -"Optional model override (`--model`, e.g. \"gpt-5.2\", \"sonnet-4-thinking\",\n   \"auto\"); validated against `cursor-agent models` and rejected on a\n   typo. Omit to use your Cursor default."New value: +"Optional model override (`--model`, e.g. \"auto\", \"gpt-5.2\",\n   \"claude-opus-4-8-high\", \"composer-2.5\"); validated against\n   `cursor-agent models` and rejected on a typo. cursor bakes the effort\n   and speed axes into the id (…-low/-high/-xhigh/-max, each with a\n   -fast twin) and also accepts a bracket form on the family base, e.g.\n   \"claude-opus-4-8[context=1m,effort=high]\". Omit to use your Cursor\n   default."
  2. Addedv0.21.0

TDQS

A4.3/5.0
Behavior3/5

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

Adds substantial context beyond the annotations: auth via Cursor login or CURSOR_API_KEY, a warning that the sandbox is agent-enforced rather than an OS boundary, and that this is an autonomous agent requiring trusted prompts. Annotations already declare readOnlyHint=false and openWorldHint=true, so the marginal gain is real but narrower than on an unannotated tool. Timeout, model validation and streaming behavior are already in the schema.

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?

Dense and front-loaded, leading with the differentiator before mechanics; every clause carries information. The stacked parentheticals and inline code tokens make it slightly harder to scan than a clean sentence-per-idea layout.

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

Completeness5/5

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

For a six-parameter agentic tool with annotations, full schema coverage and an output schema, the description covers what schema cannot: auth, sandbox fidelity, follow-up path and trust requirements. Nothing an agent needs in order to call it correctly is missing.

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

Parameters5/5

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

Schema coverage is 100%, and the description still adds meaning: it explains that workspace is effectively required for repo visibility even though it is optional, names the model families available, and characterizes the sandbox modes ('read-only = ask mode, NOT an OS boundary'). That is genuine interpretation beyond the field text.

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?

States a specific verb (ask) and resource (Cursor), scopes it to a NEW chat, and gives the identifying implementation detail (`cursor-agent -p`). The 'NEW chat' clause and 'Follow up with cursor_continue' immediately distinguish it from its sibling.

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?

Gives clear selection context ('Best for agentic coding on a Cursor plan') and routes follow-ups to cursor_continue, plus a prerequisite (pass workspace or the sub-agent sees no repo) and a restriction ('trusted prompts only'). It stops short of explicitly contrasting against other agent backends like claude_ask or codex_ask, but the guidance present is concrete and actionable.

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