[BUG] todo items are not split by \n

[feanor12]

Problem (one or two sentences)

If the model generates todo items it ignores \n in the text and treats it as a single item.

Context (who is affected and when)

vs code - general useage

Reproduction steps

VS Code Extension on Windows.
Ask for some task generation.
-> Only get a single todo item is generated but the answer shows \n characters that should split the items.

Expected result

Multiple todo items

Actual result

Single todo item

Variations tried (optional)

No response

App Version

3.51.1

API Provider (optional)

LiteLLM

Model Used (optional)

qwen3-coder-next

Roo Code Task Links (optional)

No response

Relevant logs or errors (optional)

[roomote-v0]

I've analyzed this issue and here's my proposed implementation plan:

Root cause: The parseMarkdownChecklist function in src/core/tools/UpdateTodoListTool.ts splits the input string on actual newline characters (\r?\n). However, some models/providers (like LiteLLM with qwen3-coder-next) send the todos parameter with literal escape sequences -- the two-character sequence \n (backslash + n) rather than actual newline characters. This means all checklist items end up on a single "line" and only the first item (or none) is parsed.

Proposed fix:

1. In parseMarkdownChecklist() (src/core/tools/UpdateTodoListTool.ts), normalize literal \n and \r\n escape sequences into actual newline characters before splitting. This is a one-line addition before the existing .split(/\r?\n/) call.

2. Add test cases to src/core/tools/__tests__/updateTodoListTool.spec.ts to cover the scenario where the input contains literal \n escape sequences instead of actual newlines.

This is a minimal, targeted fix that handles the mismatch between what models send and what the parser expects, without affecting any other behavior.

Please mention @roomote in a comment saying that you approve this plan, and I'll proceed with the implementation.

[feanor12]

@roomote sounds good.

[roomote-v0]

I've seen your approval and will proceed with the implementation. I'll normalize literal \n escape sequences in parseMarkdownChecklist() and add test coverage. Will open a PR shortly.

[roomote-v0]

Opened a draft PR with my attempt at addressing this: #11978. Feedback welcome!