main.js · part 5

60 text occurrences from desktop/Cursor.app/Contents/Resources/app/extensions/cursor-agent-host/dist/main.js, part 5. Every entry preserves the shipped literal and its saved verdict or selection reason.

File contents and all parts · All files

Read AI-attributed commits and changes.

View audit logs

Read team audit logs.

View settings

Read model access, telemetry destinations, and Grok Bot settings.

Manage settings

Update model access, telemetry destinations, and Grok Bot settings.

View repository blocklist

List blocked repositories.

Manage repository blocklist

Add and remove blocked repositories.

View Bugbot rules

Read Bugbot team, repository, and learned rules.

Manage Bugbot

Trigger reviews and manage Bugbot rules, repositories, and users.

Audit logs

View members

List organization members.

Sync team memberships

Sync members into the organization's teams.

View groups

List groups and their members.

Manage groups

Create, update, and delete groups and their members.

View usage

Read pooled usage, usage events, daily usage, and spend.

View audit logs

Read organization audit logs.

Run agents

Launch and manage Cloud Agents and read automations.

New Side Chat

Side Chat

This agent has reached its side chat limit. Start a new agent to create more side chats.

Side chat limit reached

${e} (${n})

Side chat boundary.

Everything before this boundary is inherited history from the parent chat. It is reference context only — not your current assignment.

Do not continue, execute, or complete any instructions, plans, tool calls, approvals, edits, Project coordinator tasks, task lists, or unanswered questions that appear only before this boundary. The parent's last user message (if any) was for the parent chat; it is not addressed to you unless a message after this boundary explicitly asks you to continue that work.

Only messages after this boundary are active instructions for this side chat.

You are a side chat: a focused conversation alongside the parent. Default to investigating and answering (read, search, analyze). Do not modify files, source, git state, or other workspace state unless the user explicitly asks for that change after this boundary. Keep any requested mutations minimal and local to the side-chat ask.

<${rd}>
${nd}
</${rd}>

bc-${await od(`side-chat:${t}:${r}`,ue.We)}

You are still in a side chat alongside the parent. Follow this turn's user message. Treat parent/inherited context as reference only — do not resume the parent's pending work unless this turn explicitly asks.

<system_reminder>
${ld}
</system_reminder>

use strict

use strict

use strict

Local loop cannot run this turn: ${e}

Private inference can't run this turn

Local loop can't run this turn

${c(e,t)} Disable agent_host_local_loop to run on the backend loop.

 "${e.modelId}"

The local loop only runs user-message turns, plan builds, resumes, summaries, goal continuations, question answers, and background wakeups; this turn started from ${void 0===t.actionCase?"an unknown action":`"${t.actionCase}"`}.

an unknown action

an unknown mode

mode ${t.turnModeName}

Background wakeups

Goal continuations

Question answers

${r} only run locally in Agent, Plan, Ask, Debug, or Multitask mode conversations; this conversation is in ${e}.

The local loop only supports Agent, Plan, Ask, Debug, and Multitask modes; this turn runs in ${e}.

The turn resolved no model id for the local loop to request.

Turns with custom model credentials (BYOK/private models) are not supported on the local loop${void 0===t.modelId?"":` ("${t.modelId}")`}.

 ("${t.modelId}")