go-llm

Go to file

Steve Dudenhoeffer e5a046a70b Handle execution errors by appending them to the result.

Previously, execution errors were only returned in the refusal field. This update appends errors to the result field if present, ensuring they are included in the tool's output. This change improves visibility and clarity for error reporting.

2025-03-17 23:41:48 -04:00

schema

initial commit of untested function stuff

2024-11-08 20:53:12 -05:00

anthropic.go

Fix role setting for assistant-sent images in Anthropic API

2025-01-09 01:18:11 -05:00

context.go

Refactor response handling for clarity and consistency.

2025-03-17 00:18:32 -04:00

error.go

initial commit of untested function stuff

2024-11-08 20:53:12 -05:00

function.go

Refactor entire system to be more contextual so that conversation flow can be more easily managed

2025-03-16 22:38:58 -04:00

functions.go

Refactor entire system to be more contextual so that conversation flow can be more easily managed

2025-03-16 22:38:58 -04:00

go.mod

make toolbox optional

2024-12-28 20:39:57 -05:00

go.sum

Refactor entire system to be more contextual so that conversation flow can be more easily managed

2025-03-16 22:38:58 -04:00

google.go

Refactor Google LLM adapter to support tool schemas.

2025-01-22 23:56:20 -05:00

llm.go

Handle execution errors by appending them to the result.

2025-03-17 23:41:48 -04:00

openai.go

Refactor entire system to be more contextual so that conversation flow can be more easily managed

2025-03-16 22:38:58 -04:00

request.go

Refactor entire system to be more contextual so that conversation flow can be more easily managed

2025-03-16 22:38:58 -04:00

response.go

Refactor response handling for clarity and consistency.

2025-03-17 00:18:32 -04:00

toolbox.go

Refactor entire system to be more contextual so that conversation flow can be more easily managed

2025-03-16 22:38:58 -04:00