[devin-ai-integration]

Summary

Closes a coverage gap flagged by a code scan: truncate_tokens was only exercised with start=0 and a text longer than max_tokens, leaving the offset slice and the already-within-limit return untested.

Adds two tests in tests/llm/test_llm_utils.py:

• test_truncate_tokens_with_start: pins the exact window tokens[start:start+max_tokens] (start=2, max_tokens=3 -> " a simple test"), so an off-by-one at either boundary fails.
• test_truncate_tokens_within_limit_returns_text_unchanged: asserts the else branch returns the original string byte-for-byte (no decode round-trip), including the exact-fit boundary len(tokens) == max_tokens.

Note the third assertion documents existing behavior that start only shrinks the window it compares against: truncate_tokens(text, max_tokens=4, start=2) returns the whole text (6 tokens) rather than the offset slice, since len(tokens) - start <= max_tokens.

No production code changed.

Devin-Org: engineering

Link to Devin session: https://app.devin.ai/sessions/949458399186449f8cd7a5b706135d59
Open in Devin Desktop: https://app.devin.ai/desktop/session/949458399186449f8cd7a5b706135d59?variant=devin

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment, CI, and merge conflict monitoring