pydantic-ai

assets

Support WebSearchTool on OpenAI GPT-Live through its delegated backend (#9049 )

2026-10-02 17:29:18 -06:00

benchmarks

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

cassettes

fix: Snap gemini-3.1-flash-image thinking efforts to minimal and high on the Gemini API

2026-09-30 21:21:47 -05:00

clai2

Bump genai-prices to 0.1.10 and keep tests passing on rich 15, dbos 3 and e2b 2.55 (#10077 )

2026-10-10 09:18:50 -04:00

durable_exec

Bump genai-prices to 0.1.10 and keep tests passing on rich 15, dbos 3 and e2b 2.55 (#10077 )

2026-10-10 09:18:50 -04:00

evals

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

example_modules

Front-page rewrite: reassurance hero, Harness in the same breath, modality tabs (#7421 )

2026-08-15 02:37:12 +00:00

ext

chore: use specific pyright: ignore rule codes and document run_stream_sync under durable execution (#6646 )

2026-07-23 16:37:36 -05:00

graph

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

harness

Bump genai-prices to 0.1.10 and keep tests passing on rich 15, dbos 3 and e2b 2.55 (#10077 )

2026-10-10 09:18:50 -04:00

models

Bump genai-prices to 0.1.10 and keep tests passing on rich 15, dbos 3 and e2b 2.55 (#10077 )

2026-10-10 09:18:50 -04:00

profiles

Add support for claude-haiku-5-5 (#9998 )

2026-10-09 10:30:11 -05:00

providers

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

realtime

Classify xAI push-to-talk held-audio send failures and clears after a request's commit in the realtime simulator (#9857 )

2026-10-09 14:06:55 -06:00

v2

Route Vertex-vs-Gemini API branches in GoogleModel by client transport instead of provider name (#7280 )

2026-08-29 23:03:28 -05:00

__init__.py

Uv workspaces and pydantic-ai-slim (#92 )

2024-11-25 21:54:23 +00:00

_inline_snapshot.py

Complete the Python 3.11 documentation and tooling cleanup (#9891 )

2026-10-06 08:43:27 -04:00

AGENTS.md

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

capability_models.py

Split tests/test_capabilities.py by topic (#7772 )

2026-08-28 23:07:28 +00:00

cassette_hooks.py

Replace pytest-recording with cassetter for cassette recording (#8893 )

2026-09-28 16:50:21 +02:00

cassette_utils.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

CLAUDE.md

Add AGENTS.md rules extracted from past PR comments (#4214 )

2026-02-05 18:58:42 +00:00

conftest.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

continuation_utils.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

cost_guards.py

Fix the CI failures on main: per-test budget, subprocess guard, and Docker Hub rate limit (#10080 )

2026-10-10 08:35:18 -04:00

import_examples.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

kw_only_walker.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

mcp_server.py

chore: use specific pyright: ignore rule codes and document run_stream_sync under durable execution (#6646 )

2026-07-23 16:37:36 -05:00

mcp_task_server.py

Allow MCPToolset clients to skip optional MCP tasks via prefer_tasks (#6851 )

2026-07-30 14:44:45 +05:30

model_lifecycle_utils.py

Context-manage models rebuilt inside durable operations (#7954 )

2026-08-31 16:39:01 -06:00

parts_from_messages.py

Expose .messages, .toolsets types in top-level pydantic_ai (#2957 )

2025-09-30 07:06:10 -06:00

temporal_utils.py

Use the CI-cached Temporal dev-server binary in the harness Temporal tests (#9650 )

2026-10-02 12:15:13 -05:00

test_ag_ui.py

Fix ID-only tool call deltas emitting a null args fragment in AG-UI and Vercel AI adapters (#9724 )

2026-10-09 11:19:10 -06:00

test_agent_output_schemas.py

Fix TextOutput JSON schema to infer the wrapped function's return annotation (#9652 )

2026-10-09 11:58:47 -06:00

test_agent.py

Allow list-form tool_choice to name output tools, and accept static lists that include one (#8565 )

2026-10-09 14:23:14 -06:00

test_builtin_tools.py

Add realtime speech-to-speech support with Agent.realtime(), with support for OpenAI, Gemini, Azure, and xAI (#6324 )

2026-08-11 08:05:41 -06:00

test_cache_prefix_invariant.py

Check cache prefix stability against the cassette cassetter already loaded (#8961 )

2026-09-28 19:54:03 +02:00

test_cache_prefix_stability.py

Replace pytest-recording with cassetter for cassette recording (#8893 )

2026-09-28 16:50:21 +02:00

test_cache_setting.py

Cache the instructions and tool definitions too when cache=True uses Anthropic's automatic caching (#10047 )

2026-10-09 11:23:04 -06:00

test_capabilities.py

Make wrap_* hooks enclose complete stage lifecycles (#7053 )

2026-10-02 17:15:49 -06:00

test_capability_combine.py

Add unified cache setting and Caching capability for cross-provider prompt caching (#7560 )

2026-10-08 09:57:56 -06:00

test_capability_deferred_loading.py

Load a capability once when load_capability is called twice for it in one response (#9397 )

2026-10-02 17:33:15 -06:00

test_capability_events.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_capability_hooks.py

Cancel the streamed model request when a parallel InputGuardrail blocks (#10003 )

2026-10-08 13:32:24 -04:00

test_capability_image_generation.py

Fix ImageGeneration under FallbackModel and dataclasses.replace, and compare Embedder by identity (#9477 )

2026-10-02 17:19:22 -05:00

test_capability_native_or_local.py

Fix ImageGeneration under FallbackModel and dataclasses.replace, and compare Embedder by identity (#9477 )

2026-10-02 17:19:22 -05:00

test_capability_node_streaming.py

Make wrap_* hooks enclose complete stage lifecycles (#7053 )

2026-10-02 17:15:49 -06:00

test_capability_output_hooks.py

Make wrap_* hooks enclose complete stage lifecycles (#7053 )

2026-10-02 17:15:49 -06:00

test_capability_output_lifecycle.py

Make wrap_* hooks enclose complete stage lifecycles (#7053 )

2026-10-02 17:15:49 -06:00

test_capability_pending_messages.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

test_capability_process_event_stream.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_capability_spec.py

Add an OpenAIDecisionsModel backend with image input (#9634 )

2026-10-09 14:05:45 -05:00

test_capability_stream_teardown.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_capability_visit_and_replace.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_cassette_hooks.py

Replace pytest-recording with cassetter for cassette recording (#8893 )

2026-09-28 16:50:21 +02:00

test_cassette_utils.py

Make cassette test helpers handle cassetter cassettes and fail when there is nothing to check (#9084 )

2026-09-29 10:23:19 +00:00

test_choices.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_cli.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_code_execution_files_vcr.py

Default AnthropicModel's max_tokens to the model's maximum output, streaming such requests behind the scenes (#9298 )

2026-09-29 23:15:04 +00:00

test_concurrency.py

Complete the Python 3.11 documentation and tooling cleanup (#9891 )

2026-10-06 08:43:27 -04:00

test_conftest.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_conversation.py

Add Conversation to carry and store a conversation between runs, accepted as conversation= by every entry point (#8337 )

2026-10-04 14:57:38 -06:00

test_cost_guards.py

Fix the CI failures on main: per-test budget, subprocess guard, and Docker Hub rate limit (#10080 )

2026-10-10 08:35:18 -04:00

test_cost.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_custom_events.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_deps.py

fix(test): set provider_name on TestModel non-streaming responses (#6066 )

2026-07-01 13:03:28 +05:30

test_direct.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

test_display.py

Support Python 3.15 (#10064 )

2026-10-09 23:30:25 -04:00

test_docs_navigation.py

Restore the pydantic-ai-harness base-install, pydantic-clai2 clipboard and docs-navigation CI checks lost in the monorepo move (#9655 )

2026-10-02 11:25:41 -05:00

test_docs_parity.py

Put Logfire and the Gateway on the paths newcomers and coding agents take (#9030 )

2026-09-29 17:16:42 -06:00

test_embeddings.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_exa.py

Deprecate the Exa search common tools in favor of the ExaSearch capability in Pydantic AI Harness (#6644 )

2026-07-21 22:41:44 -05:00

test_examples.py

Fix the CI failures on main: per-test budget, subprocess guard, and Docker Hub rate limit (#10080 )

2026-10-10 08:35:18 -04:00

test_exceptions.py

Explain the Bedrock data retention opt-in when Claude Fable 5 / 5.1 reject the default mode (#10050 )

2026-10-09 14:21:20 -06:00

test_fallback_native_factory.py

Fix ImageGeneration under FallbackModel and dataclasses.replace, and compare Embedder by identity (#9477 )

2026-10-02 17:19:22 -05:00

test_format_as_xml.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

test_function_schema.py

Add @durable_operation for capabilities and a public backend API for third-party durable execution engines (#6696 )

2026-08-28 15:49:34 -06:00

test_function_signature.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

test_gateway_model_name_suggestions.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

test_history_processor.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_http_client_defaults.py

Default Pydantic AI-created HTTP clients to the OpenAI SDK's connection pool limits, and make create_async_httpx2_client() public (#9600 )

2026-10-02 17:09:24 -06:00

test_http_client_timeouts.py

Stop a numeric request timeout from lengthening the connect and pool timeouts of Pydantic AI-created HTTP clients (#9599 )

2026-10-04 15:36:46 -06:00

test_httpx2_sdk_readiness.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_images.py

Raise ModelAPIError when a 200 response from OpenAI Responses, Anthropic, Groq or OpenAI images has a text/plain body (#10065 )

2026-10-09 14:29:38 -06:00

test_include_binary_content.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_instruction_ids.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_instrumentation_cache_health.py

Add unified cache setting and Caching capability for cross-provider prompt caching (#7560 )

2026-10-08 09:57:56 -06:00

test_instrumentation_content_redaction.py

Record each failed FallbackModel attempt with its model, timing, error and usage, and count rejected responses in RunUsage (#8745 )

2026-10-09 10:05:18 -06:00

test_json_schema.py

Handle JSON-pointer $refs and draft-07 definitions in MCP TypeScript SDK tool schemas (#9643 )

2026-10-06 15:12:07 -05:00

test_local_workspace_capability.py

Add workspaces: ctx.workspace gives tools one API for files and commands, on your machine or in a sandbox, with durable execution support (#6492 )

2026-09-28 19:08:20 +05:30

test_local_workspace.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_logfire.py

Add unified cache setting and Caching capability for cross-provider prompt caching (#7560 )

2026-10-08 09:57:56 -06:00

test_mcp.py

Apply auth to requests through MCPToolset(http_client=...) and stop closing the user-supplied client (#10051 )

2026-10-09 14:22:13 -06:00

test_messages.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_model_name_suggestions.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

test_model_request_message_persistence.py

Make wrap_* hooks enclose complete stage lifecycles (#7053 )

2026-10-02 17:15:49 -06:00

test_native_output_schema.py

feat: support template=False on PromptedOutput and NativeOutput to disable schema prompt (#4497 )

2026-03-02 14:04:36 -06:00

test_native_tool_search_vcr.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_nested_sync_agent.py

Reject Agent.run_sync() from synchronous callbacks inside agent runs (#7277 )

2026-08-19 11:09:20 +05:30

test_on_event.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_otel_context_isolation.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_parts_manager.py

Preserve stream part indexes during apply_event replay (#7138 )

2026-08-04 14:32:29 -05:00

test_prices.py

Add RequestUsage.audio_seconds so duration-billed voice models are priced (#8427 )

2026-09-25 01:14:48 +00:00

test_public_interface_contracts.py

Add Conversation to carry and store a conversation between runs, accepted as conversation= by every entry point (#8337 )

2026-10-04 14:57:38 -06:00

test_readonly_workspace.py

Add workspaces: ctx.workspace gives tools one API for files and commands, on your machine or in a sandbox, with durable execution support (#6492 )

2026-09-28 19:08:20 +05:30

test_realtime_examples.py

Handle barge-in in the session: handle_barge_in=True, interrupt(played_bytes=...), and played_audio_bytes (#7870 )

2026-09-04 11:48:17 -06:00

test_ref_sibling_wire_contract.py

Replace pytest-recording with cassetter for cassette recording (#8893 )

2026-09-28 16:50:21 +02:00

test_run_cancellation.py

Keep completed tool returns under parallel_ordered_events when a sibling raises or the run is cancelled (#10049 )

2026-10-09 11:15:59 -06:00

test_run_context_context_window.py

Add context_window to ModelProfile and context_window_used to RunContext (#4611 )

2026-09-02 16:41:05 +05:30

test_run_context_usage_limits.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_run_result_serialization.py

Publish AgentRunResult's public shape in its validation JSON schema (#9791 )

2026-10-05 09:55:24 -06:00

test_sanitize_messages.py

Add workspaces: ctx.workspace gives tools one API for files and commands, on your machine or in a sandbox, with durable execution support (#6492 )

2026-09-28 19:08:20 +05:30

test_settings.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_skill_examples.py

Complete the Python 3.11 documentation and tooling cleanup (#9891 )

2026-10-06 08:43:27 -04:00

test_spec.py

Add Capability abstraction, AgentSpec, Hooks, unified thinking setting, per-run toolset isolation, builtin tool fallback to local tools (#4640 )

2026-03-24 15:45:00 -06:00

test_ssrf.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_streaming_errors.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

test_streaming.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_sync_stream_loop_affinity.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_tavily.py

Rehaul Tavily search tool (#4158 )

2026-03-06 15:10:46 -06:00

test_template.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_tenacity.py

Catch overflowing Retry-After HTTP dates (#10023 )

2026-10-09 09:34:14 -06:00

test_thinking_native_tools.py

Replace pytest-recording with cassetter for cassette recording (#8893 )

2026-09-28 16:50:21 +02:00

test_thinking_part.py

Improve CI performance with Depot runners and inline_snapshot bypass (#4346 )

2026-02-17 23:22:19 +00:00

test_thinking_wire_contract.py

fix: Snap gemini-3.1-flash-image thinking efforts to minimal and high on the Gemini API

2026-09-30 21:21:47 -05:00

test_thinking.py

Don't force output tools on Anthropic requests that think, so Claude keeps thinking with a structured output_type (#8812 )

2026-09-28 16:39:07 -06:00

test_tool_availability_portability.py

Keep a first-position ToolAvailabilityDeltaPart announcement out of the standing system prompt (#9413 )

2026-10-02 17:33:44 -06:00

test_tool_availability.py

Re-resolve tools when capability availability changes mid-step so prepare_tools still governs them (#8071 )

2026-09-16 19:50:58 -06:00

test_tool_failed_wire.py

Bump anyio dev floor to 4.11 and run async tests via anyio_mode = "auto" (#8874 )

2026-09-28 08:45:29 +00:00

test_tool_search.py

Index the default keyword tool search corpus (#9958 )

2026-10-08 11:03:58 -04:00

test_tools.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

test_toolsets.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_transcript_repair.py

Fix history repair for a tool_call_id repeated within one response (#9819 )

2026-10-09 09:00:20 -06:00

test_ui_web.py

Return a 422 instead of a 500 from Agent.to_web() for invalid chat requests (#10042 )

2026-10-09 09:42:53 -06:00

test_ui.py

Add workspaces: ctx.workspace gives tools one API for files and commands, on your machine or in a sandbox, with durable execution support (#6492 )

2026-09-28 19:08:20 +05:30

test_usage_attribution.py

Report each agent run's own usage on its span (#8456 )

2026-09-17 20:26:32 -06:00

test_usage_extraction.py

Warn on unexpected RequestUsage.extract() failures (#8564 )

2026-09-29 16:37:36 -05:00

test_usage_limits.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

test_utils.py

Fix merge_json_schema_defs corrupting branch schemas on transitive def-name collisions (#9735 )

2026-10-09 11:26:09 -06:00

test_uuid.py

Make run ID time-sortable; add agent name and run ID span attrs to all agent run child spans (#4876 )

2026-03-30 16:37:58 -06:00

test_validation_context.py

fix(test): set provider_name on TestModel non-streaming responses (#6066 )

2026-07-01 13:03:28 +05:30

test_vercel_ai.py

Fix ID-only tool call deltas emitting a null args fragment in AG-UI and Vercel AI adapters (#9724 )

2026-10-09 11:19:10 -06:00

test_web_fetch.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_workspace_conformance.py

Speed up CI by fixing the slowest tests, balancing shards by measured duration, and guarding test cost (#10054 )

2026-10-09 14:05:31 -06:00

test_workspace.py

Keep the live workspace on a settled StreamedRunResult.result (#10048 )

2026-10-09 11:18:24 -06:00

typed_agent.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

typed_deps.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

typed_graph.py

Target py311 in Ruff and modernize compatible syntax (#9854 )

2026-10-06 07:07:09 -04:00

typed_output.py

Accept Annotated, Literal and union output_types as PEP 747 TypeForms, and bump pyright to 1.1.414 (#8690 )

2026-09-23 17:23:06 -06:00

workspace_fakes.py

Add workspaces: ctx.workspace gives tools one API for files and commands, on your machine or in a sandbox, with durable execution support (#6492 )

2026-09-28 19:08:20 +05:30