fix: remove overly conservative context length check for @file (#12165)

[ChenXin-2009]

Description

Fixes #12165

The @file\ context provider had an \isItemTooBig\ check that incorrectly rejected files larger than ~16kB when using Ollama with \model: AUTODETECT. This happened because AUTODETECT skips \ensureModelInfo()\ (see \core/llm/llms/Ollama.ts:178), so \contextLength\ defaults to 8192 tokens instead of the actual model context size.

The formula used: \ okens > contextLength - maxTokens\ → \ okens > 8192 - 4096 = 4096\ rejected files that produce >4096 tokens (~16kB of text).

Fix

Removed the frontend-side \isItemTooBig\ check entirely, making @file\ consistent with @CurrentFile\ and all other context providers. The core \compileChatMessages\ function already handles context overflow gracefully.

Changes

• \gui/src/components/mainInput/AtMentionDropdown/index.tsx: Removed \isItemTooBig, \handleItemTooBig, \ormatFileSize\ dead code (~107 lines) and unused imports. All non-action/non-submenu/non-query items now go through \props.command()\ uniformly.

---

Summary by cubic

Removed the overly conservative context-length check for @file mentions that blocked valid files (notably with Ollama model: AUTODETECT). @file now behaves like @CurrentFile, with overflow handled by core logic.

• Bug Fixes
  • Removed the frontend isItemTooBig check in AtMentionDropdown; all non-action items now go through props.command(...).
  • Cleaned up dead code and unused imports.

^{Written for commit fff283d. Summary will update on new commits.}

[github-actions]

All contributors have signed the CLA ✍️ ✅
_{Posted by the CLA Assistant Lite bot.}

[cubic-dev-ai]

No issues found across 1 file

_{Re-trigger cubic}

[ChenXin-2009]

I have read the CLA Document and I hereby sign the CLA