…n chat responses

- Add tokenUsage field to IResultMetadata interface
- Update defaultIntentRequestHandler to store token info in metadata instead of inline markdown
- Enhance chatParticipantRequestHandler to display token usage in response details footer
- Token usage now appears alongside model info (e.g., 'GPT-4o • 0x • 2822/123901 tokens')
- Provides cleaner, structured approach using VS Code's ChatResult.metadata system

…y.tsx

- Revert conversationHistory.tsx to main branch version
- Keep only essential token usage functionality changes
- Removes unnecessary JSX formatting and null coalescing changes

* Fix thinking tool
#259934

* Update

* send thinking tokens progress to chat ui

* fix merge conflict:

* pass in id and metadata as well

* wip

* make it work

* error handling and retry on an invalid stateful marker

* Basic statefulMarker round-trip

* add createRequestBody on endpoints

* adopt createRequestBody in responses

* adopt generalized audience

Adopts microsoft/vscode#259273

* Slice messages array based on stateful marker data

* Send statefulMarker via LanguageModelDataPart

* Stateful marker container value is not JSON

* Properly ignore stateful marker when requested

* add setting for byok responses api, log view

* encode model with stateful marker

* Add setting to toggle responses api truncation vs our summarization

* Fix AutoChatEndpoint after merge

* vitest

* fixup

* Try enabling response for azure

---------

Co-authored-by: Connor Peet <[email protected]>
Co-authored-by: Logan Ramos <[email protected]>

* Add command to save current chat log as Markdown

This commit introduces a new command to save the currently opened chat log to a file in Markdown format. It includes a dialog for selecting the save location and handles errors during the save process.

| File                                      | Changes Made                                      |
|-------------------------------------------|--------------------------------------------------|
| src/extension/log/vscode-node/requestLogTree.ts | Added `saveCurrentMarkdownCommand` and its implementation |

* Enhance save current chat log command in requestLogTree

- Improved command to save the currently opened chat log by accepting a resource from menu invocation.
- Added fallback to the active editor's document if no resource is provided.
- Updated logic to read text from the virtual document URI explicitly before saving.

| File                                   | Changes Made                                      |
|----------------------------------------|--------------------------------------------------|
| src/extension/log/vscode-node/requestLogTree.ts | Enhanced command functionality and error handling |

* Refactor save dialog to use .md extension for Markdown files

- Changed default filename extension from .copilotmd to .md
- Updated save dialog filters to reflect the new .md extension

| File                                   | Changes                          |
|----------------------------------------|----------------------------------|
| src/extension/log/vscode-node/requestLogTree.ts | Updated filename and filters for save dialog |

Closes microsoft/vscode-internalbacklog#5723

* Update model selection checks

* nit: undo unrelated change

Closes microsoft/vscode#250017

Refs microsoft/vscode#257379

I hypothesize that what's happening is the model tries to make an edit
with an invalid `oldString`. We then go to healing and the healing
process incorrectly causes the `newString` to no-op. This makes a
couple changes:

- Explicitly throw if oldString===newString, which we will get in telemetry
- Only do the `unescapeStringForGeminiBug` for Gemini, not Claude. Claude
  may not have this issue and is generally really good at self-correcting,
  I strongly suspect this code path is where things go awry.

* Enable enableRetryAfterFilteredResponse in Insiders

* Fix

* add length telemetry event with log message

* convert function arguments to length

* add completionId

* remove tokens

* add modelCallId

* fix modelCallId

* fix modelCallId

* move modelCallId from inside messages

* add type guard for 'tool_calls'

* add curly brackets

* remove duplicate noChoice

* use isOutput instead of guessing input/output

* fix usage issue

* change logging to debug mode

---------

Co-authored-by: Yevhen Mohylevskyy <[email protected]>
Co-authored-by: Yevhen Mohylevskyy <[email protected]>

* Delete knowledge bases

* Also delete SSO stuff
Which I am pretty confident is only used for knowledge bases

* Use lineNumberStyle for FileVariable

* vitest

* update test files with no baseline...

* Update baseline scores for toolCalling tests to reflect recent changes

* tests

* Update baseline

---------

Co-authored-by: Ulugbek Abdullaev <[email protected]>

* Remove some unused settings

* Another

* rely more on observable document due to notebooks

* set debounceDelayMs to 0

* Fix backtick in search query breaking tool result codespan

* Fix

Closes microsoft/vscode#257379

* separate added/removed lines

* leave undefined

…#526)

Will need to fix these telemetry events for this API later

Closes microsoft/vscode#259690

* Update getErrors (GetErrors / problems) tool

Update this tool so it can gather all errors.

* Updated model description so agent knows the tool can gather all errors
* Updated implementation such that if no URL is given, the tool returns all
errors.

* Update tool definition

Made the 'filePaths' argument optional, and updated description
to indicate how the argument should be used.

* Updates from review

* Made filePaths on IGetErrorParams optional, updated implementation of
invoke accordingly.
* Updated prepareInvoke to handle optional argument properly as
well.

* Tweaks

---------

Co-authored-by: Rob Lourens <[email protected]>

Took a long time, this started as a PR on the old vscode-copilot repo

* Add GPT-4.1 agent prompt and enhance tool capabilities

Implemented the GPT-4.1 specific agent prompt to improve multi-step task execution and problem-solving. Updated the DefaultAgentPrompt to conditionally render the new GPT41AgentPrompt based on the model family. Enhanced the tool capabilities detection function to streamline tool usage and added a new tool name for managing todo lists.

* Add setting for new prompt, tweaks

* Cleanup

---------

Co-authored-by: Rob Lourens <[email protected]>

* Initial plan

* Add NES requests filter implementation

Co-authored-by: roblourens <[email protected]>

* Refine NES request identification logic to avoid false positives

Co-authored-by: roblourens <[email protected]>

* Simplify

---------

Co-authored-by: copilot-swe-agent[bot] <[email protected]>
Co-authored-by: roblourens <[email protected]>

* Add some simple NES tests for alt format of notebooks

* Updates

* Updates

No model actually uses this config currently, but fix the instructions nevertheless

Closes microsoft/vscode#258888

* add CoreManageTodoList tool to packagejson tools for testing

* fix: update references from CoreTodoListTool to CoreManageTodoList

* Move terminal selection/command tools to core

Part of microsoft/vscode#259260

* Update baseline

…uest Log tree

…th comma separators