agentjido / req_llm / 59f7c6f6b50d40a11a5550aab27255b3024c7011

Builds Branch Commit Type Ran Committer Via Coverage
59f7c6f6b50d40a11a5550aab27255b3024c7011 main Refactor context and message handling for tool calls (#98) * Refactor context and message handling for tool calls - Updated `ReqLLM.Context` to replace `push_user`, `push_assistant`, and `push_system` with `append` and `prepend` methods for bett... push web-flow github
51.23
9d88c6edb6ffc04164d98446bcefb09536925156-PR-100 refactor/xai-structured-objects Enhance XAI provider support and update fixtures - Added new structured output test cases for the XAI provider, improving coverage for various scenarios. - Introduced multiple new fixture files for XAI models, including basic, multi-tool, and str... Pull #100 mikehostetler github
51.95
37350553ccef71900d72498cba84aa5e42954e33-PR-66 bedrock-clean Add extended thinking and multi-turn tool calling support - Add pre_validate_options/3 to translate reasoning_effort to budget_tokens - Support Claude extended thinking via additionalModelRequestFields (Converse) and thinking parameter (native) -... Pull #66 neilberkman github
53.57
664b4e288eb6e63d496872bd79a0508852c65cd0-PR-99 chore/2025-10-14-update-fixtures Massive test fixture update Pull #99 mikehostetler github pending completion   set done
9e02a5713331885b030771b7caf2759f7ccd926a-PR-98 refactor/context-tools Refactor assistant message handling in ReqLLM.Context - Updated `ReqLLM.Context` to streamline the creation of assistant messages, allowing for optional tool calls and metadata. - Replaced the deprecated `assistant_with_tools` function with a mor... Pull #98 mikehostetler github
51.19
2af1c5016f12a5c077f1484ade774cf2197330ff-PR-98 refactor/context-tools Add example scripts for context reuse, cross-model context, and reasoning tokens - Introduced new example scripts demonstrating various functionalities of the ReqLLM library, including: - `context_reuse.exs`: Shows how the same context can be p... Pull #98 mikehostetler github
51.83
98ddb061b9d839e52ea022c24c28893781d527cb-PR-98 refactor/context-tools Refactor context and message handling for tool calls - Updated `ReqLLM.Context` to replace `push_user`, `push_assistant`, and `push_system` with `append` and `prepend` methods for better clarity and consistency. - Introduced `ReqLLM.ToolCall` str... Pull #98 mikehostetler github
51.71
238938f570ca339103ec9eff557676f7cef21770 main Resurrect: Enhance model compatibility task and update provider implementations (#88) * Implement object generation support check and enhance StreamServer functionality - Added `supports_object_generation?/1` function in `ReqLLM.Capability` to d... push web-flow github
52.0
486e0f3e50cfe8ff890c3d9602172a856792daad-PR-88 chore/object-fixtures-resurrected Add example scripts for embeddings, object generation, and multimodal analysis - Introduced multiple example scripts demonstrating various functionalities of the ReqLLM library, including: - `embeddings_single.exs`: Generates embeddings for a s... Pull #88 mikehostetler github
52.1
50e5599fc8058c36e223672e89bd287e948ca840-PR-88 chore/object-fixtures-resurrected Clean up dbug methods Pull #88 mikehostetler github
51.95
829a3739d3a055dbc18602d736648de7fb641248 main fix(google): implement structured output using native responseSchema (#89) Implementation follows OpenAI's json_schema pattern where the model returns JSON as text content, which is then parsed and set in the response object field. Key changes: ... push web-flow github
48.93
387b4e6e52ea1d55e39083ddcd51abda73ac9d73-PR-83 fix-google Allow stream to return response body on error responses Pull #83 scrogson github
49.28
85df690ba8bcffaacde41f6f48320cc255ae120a main fix: Respect max_tokens from Model.new/3 across all providers (#95) * Add model options extraction to user options processing * Add tests for model options extraction in Options.process/4 * Improve max_tokens extraction logic to ensure positive... push web-flow github
48.84
799236d071d64072b20e90dd25d3bc0975dd4572-PR-95 put-max-tokens-model-options Add test to ensure max_tokens: 0 is not extracted for embedding models Pull #95 lucas-stellet github
48.77
e6d5bc142f2879d6363b3322e3c8c03a11d0abe0 main Fix Anthropic provider tool result encoding for multi-turn conversations (#94) Resolves critical bugs preventing tool calling from working with Anthropic models: 1. Transform :tool role to :user - Anthropic API only accepts "user" or "assistant"... push web-flow github
48.87