3
49%
main: 49%

Ran 18 Nov 2025 01:42AM UTC

Files 82

Run time 4s

Badge

Embed ▾

Committed 18 Nov 2025 01:41AM UTC coverage: 50.14% (+0.08%) from 50.056%

Job # 20e498f7947cf72d00b9b1f324db8f49b6d34a14.3

Build Type

push

github

Committed by

web-flow

Commit Message

Fix tool_budget_for pattern match regression from LLMDB integration (#221)

The test helper tool_budget_for/1 was using an obsolete pattern match
that expected {:ok, {provider, id, model}} but ReqLLM.model/1 now returns
{:ok, model} after the LLMDB integration in v1.1.0.

This caused the pattern match to fail and always return the default 150
tokens instead of calculating the correct budget based on model.limits.

Additionally fixed the field name from model.limit (singular) to
model.limits (plural) to match the LLMDB.Model struct.

The regression primarily affected reasoning-enabled models like Gemini
2.5 Pro, where 150 tokens was insufficient for the model's extended
reasoning process, causing tool call tests to fail with MAX_TOKENS
errors.

Run Details

3588 of 7156 relevant lines covered (50.14%)

14.69 hits per line

agentjido / req_llm / 20e498f7947cf72d00b9b1f324db8f49b6d34a14 / 3
49%
main: 49%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Source Files on job 20e498f7947cf72d00b9b1f324db8f49b6d34a14.3

agentjido / req_llm / 20e498f7947cf72d00b9b1f324db8f49b6d34a14 / 3 49% main: 49%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Source Files on job 20e498f7947cf72d00b9b1f324db8f49b6d34a14.3

agentjido / req_llm / 20e498f7947cf72d00b9b1f324db8f49b6d34a14 / 3
49%
main: 49%

README BADGES
x