workbench.desktop.main.js · part 264

60 text occurrences from desktop/Cursor.app/Contents/Resources/app/out/vs/workbench/workbench.desktop.main.js, part 264. Every entry preserves the shipped literal and its saved verdict or selection reason.

File contents and all parts · All files

CloudAgentStream: skipping stale initialState snapshot

Plan written to ${vgv(s.additionalData)||"unknown plan file"}:
${a}

unknown plan file

No output

use strict

use strict

CloudSubagentTranscriptProjection: failed to project cloud subagent update

Unsupported tool type

use strict

use strict

use strict

[ConversationClassification] Manually triggering classification for recent conversations...

[ConversationClassification] Classification complete. Total conversations scored: ${b}

Classification triggered

[ConversationClassification] Error during manual classification:

[ConversationClassification] Error during classification polling:

[ConversationClassification] Failed to store summary locally:

[ConversationClassification] Fire-and-forget summarization failed (non-fatal):

[ConversationClassification] Error parsing stored scored conversations:

[ConversationClassification] Error saving scored conversations:

use strict

Checking status...

Browser automation disabled by admin

Unable to verify browser policy

Connected to Browser Tab

Browser automation disabled

use strict

use strict

use strict

use strict

use strict

Failed to cache admin settings:

Failed to fetch team admin settings:

Failed to parse cached admin settings:

<redacted: ErrorNoTelemetry>

[AutorunSettings] Failed to fetch admin settings (network error, no cache available):

[AutorunSettings] Admin settings temporarily unavailable but were previously enforced - keeping existing restrictions

Failed to update workspace trust setting:

Invalid allowedExtensions format: expected object, got ${Array.isArray(L)?"array":typeof L}

Invalid allowedExtensions format

Enable Run Everything?

Enable Run Everything

Use Sandbox instead

Use Allowlist instead

Leave Ask Every Time?

use strict

Commands run in a protected sandbox that limits access to your files, network, and git. You can allowlist specific commands to run with full access when needed. Be cautious of potential prompt injection risks.

Only commands you've added to your allowlist will run automatically. Other commands will require user approval. Be cautious of potential prompt injection risks.

This allows the agent to execute any tool or shell command without approval. A prompt injection or a malicious tool could delete files or exfiltrate secrets from your machine. We recommend using Sandbox.

This allows the agent to execute any tool or shell command without approval. A prompt injection or a malicious tool could delete files or exfiltrate secrets from your machine. We recommend using Allowlist.

Ask Every Time will disappear from this menu after you switch away from it. Allowlist auto-runs tools that are already in your allowlist, including MCP tools. If you want every shell command to keep asking for approval, keep your shell allowlist empty. With an empty shell allowlist, shell commands still ask every time and Cursor skips shell allowlist parsing.

use strict

Whether the browser tab is enabled

use strict

use strict

use strict

use strict

use strict

Unknown MCP server group: ${e}

Selection for MCP server group ${e} must include at least one group member