Repo sync for protected branch#359
Open
learn-build-service-prod[bot] wants to merge 58 commits intomainfrom
Open
Repo sync for protected branch#359learn-build-service-prod[bot] wants to merge 58 commits intomainfrom
learn-build-service-prod[bot] wants to merge 58 commits intomainfrom
Conversation
* Added page for GitHub Copilot SDK * Small fix
Merge main to live
* Add an overview for all integrations * Fix warnings
Merge from main to live
Renamed Github to GitHub
* Update magentic orchestration doc * Add important node on group chat custom orchestrator
update to AIFunction to FunctionTool
Fix for Purview link
Fixed integrations link
Merge main to live
* Added docs for Claude Agent SDK * Small update * Addressed comments
Merge main to live
The [!IMPORTANT] block in executors.md already documents the new [MessageHandler] attribute pattern. Removed duplicate [!NOTE] blocks from other files that were cluttering the documentation.
Rename GetNewSession to CreateSession
Merge changes from main to live
Python: main to live
Obsolete reflecting executor
Merging updates about ReflectingExecutor being obsoleted.
* docs: Add Python 2026 breaking changes guide and update docs - Create python-2026-breaking-updates.md with all breaking changes since 2026 - Update context_providers → context_provider in memory docs - Update source_executor_id → executor_id in workflow docs - Update create_agent() → as_agent() references - Fix middleware examples to use list format * chore: Add copilot instructions * Add agent-framework PR merge agent for docs and breaking changes * updated topics * fix syntax * Add scheduled workflow for agent-framework docs sync * Update workflow to use Copilot coding agent via issue assignment * Use previous workflow run time instead of explicit dates * Address PR review comments - Use PR labels (breaking-change) instead of title prefix to identify breaking changes - Update example to use real PR #3413 (AIFunction rename) instead of fictional #3500 - Note: Content restructuring change is already documented under python-1.0.0b260123 (line 164) * Check both PR labels and title for breaking changes * Expand Content types migration guide with full reference table (#3252) * Add Annotation and TextSpanRegion migration guide (#3252) * disabled scheduled run * fix source event and added note on transient nature
…kage, and event model changes (#852) * docs: Update workflows docs for state API, orchestrations package, and event model changes * Revert changes doc import
* Update docs to show support for decorator types * Fix sample code
Python: Renamed AgentProtocol to SupportsAgentRun
…changes (#856) * Updates to docs based on latest event type * Fix typo * Update docs based on fluent methods changes * Updates * Add changes for removing factories
Merge main to live
Adjust docs for Magentic c# (#864)
Update overview page with "Why Microsoft Agent Framework?"
This reverts commit 595d651.
Merge main to live
.Net: Use default azure credentials
* Renamed AgentProtocol to AgentLike * Renamed AgentLike to SupportsAgentRun * restructure: create new docs structure with TOC and move old files to _to_delete New structure: - get-started/ (6 progressive tutorial pages) - agents/ (overview + tools, middleware, conversations, providers subsections) - workflows/ (11 workflow pattern pages) - integrations/ (a2a, ag-ui, azure-functions, openai-endpoints, purview, m365) - New TOC.yml reflecting complete restructure Existing content moved to new locations where possible. Placeholder pages created for new sections (will be populated with :::code refs). Old tutorials/ and user-guide/ moved to _to_delete/ for review. DevUI, support, migration-guide, api-docs kept as-is. * docs: add :::code references to get-started pages and fix links - All 6 get-started pages now use :::code language source id highlight - Fixed pip install from agent-framework[openai] to agent-framework --pre - Fixed broken links in overview/index.md (user-guide → new structure) - Added next-step navigation and go-deeper lateral links to all pages * feat: add streaming snippet refs to your-first-agent.md for Python * docs: add AGENTS.md with structure & design choices for AI/human maintainers * fix: repair 120+ broken links, update TOC with migration/support sub-TOCs, add human-in-the-loop - Fixed all broken cross-references from old user-guide/tutorials structure to new layout - Main TOC: added from-autogen and from-semantic-kernel sub-TOCs under Migration Guide - Main TOC: linked support/TOC.yml for full support section (FAQ, troubleshooting, upgrade) - Added workflows/human-in-the-loop.md (was in samples but missing from docs) - support/TOC.yml: added FAQ and Troubleshooting entries * refactor: rewrite overview page — code-first, half the length - Opens with runnable code snippet (create agent + run) before any explanation - Zone pivots for C#/Python with :::code refs to sample files - Replaced verbose sections with concise table (agents vs workflows, when to use) - Cut from 160 lines to ~80 lines - Added nextstepaction and Go deeper navigation * docs: inline code, add skills files, update AGENTS.md - Replace all :::code references with inline code blocks - .NET code sections have TODO placeholders - Add .github/skills/code-snippets.md (:::code directive guide) - Add .github/skills/sample-structure.md (progressive complexity principles) - Update AGENTS.md to reflect current inline approach * docs: switch to AzureOpenAI Foundry provider, fix duplicated code blocks - Update get-started pages and overview to use AzureOpenAIResponsesClient with Azure AI Foundry project endpoint - Fix memory.md and hosting.md where full file was pasted multiple times (snippet IDs not found) — replaced with clean focused code blocks * cleanup: remove TODO/source comments, add .NET placeholders, add skills frontmatter - Replace empty .NET code blocks with 'coming soon' notes - Add YAML frontmatter to code-snippets skill - Rewrite sample-structure skill to be conceptual-only (language-specific details belong in code repo skills files) * feat: add .NET code to all docs pages, fix Python duplication - Write real .NET code for all get-started pages (overview, first-agent, add-tools, multi-turn, memory, workflows, hosting) - Write .NET code for 13 concept pages (middleware, providers, tools, workflows) - Replace 'complete example' placeholders on 17 pages with sample links - Fix duplicated Python content in workflows.md - All 60 placeholder notes resolved — zero remaining * fix: apply upstream renames (ChatAgent→Agent, ChatMessage→Message) to new structure Upstream commits renamed Python classes: - ChatAgent → Agent (from agent_framework import Agent) - ChatMessage → Message (from agent_framework import Message) - call_next no longer takes context parameter Applied these renames to all active docs pages. * fix: resolve docs build errors and warnings * fix: resolve invalid-zone error in local-mcp-tools.md Separate zone pivot attribute from inline text to fix the only build error (invalid-zone at line 132). * feat: expand AG-UI integration into full subfolder with all content Move AG-UI from single page to subfolder with getting-started, backend/frontend tool rendering, security, HiTL, state management, and testing pages. Update TOC to reference subfolder. * fix: update broken links after AG-UI subfolder move - get-started/hosting.md: ag-ui.md → ag-ui/index.md - integrations/overview.md: ag-ui.md → ag-ui/index.md - ag-ui/backend-tool-rendering.md: old tutorials path → agents/tools/ - ag-ui/human-in-the-loop.md: old tutorials path → agents/tools/ * fix: complete provider support matrix for tools Split OpenAI/Azure OpenAI into Chat Completion, Responses, and Assistants columns (Azure mirrors OpenAI). Add links from rows to tool pages and columns to provider pages. Add Image Generation row for Python. Fix tool support accuracy per codebase analysis. * fix: update hosted MCP tool code to use client.get_mcp_tool() Replace deprecated HostedMCPTool(...) with client.get_mcp_tool(...). Fix parameter names: server_label→name, server_url→url, require_approval→approval_mode. * feat: rewrite OpenAI provider page to cover all three client types Split into Chat Completion, Responses, and Assistants sections with accurate tool support per client. Added get_*_tool() method examples for Responses client. Updated title and description. * feat: add Python A2A integration with zone pivots Add Python section covering A2AAgent usage: connecting to remote agents, streaming, long-running tasks with continuation tokens, and authentication. Add zone_pivot_groups for language switching. * feat: add Python zone to OpenAI-compatible endpoints page Add Python section showing how to connect to any OpenAI-compatible endpoint using base_url on OpenAIChatClient/OpenAIResponsesClient. Covers Ollama, LM Studio, vLLM, and hosted agents. Add zone pivots. * fix: add agent-framework-overview.md to TOC * removed _to_delete after merge * fix middleware syntax * add skills * add redirection skill * fix: resolve build warnings and add redirections for deleted pages - Add 91 redirect entries for deleted tutorials/, user-guide/, and renamed files - Remove stale integrations/index.md redirect (file exists now) - Exclude AGENTS.md from docfx build (not a published page) - Fix broken image paths in azure-functions.md and orchestration pages - Fix overview.md -> index.md references across all AF content - Fix broken cross-refs (multi-turn-conversation, agent-types, using-agents) - Consolidate overview/index.md content (manual update) - Remove deleted overview/agent-framework-overview.md from TOC * fix: replace A2A hosting with Azure Functions in getting-started docs - hosting.md: Python section now shows AgentFunctionApp pattern - hosting.md: Add pip install agent-framework-azurefunctions requirement - hosting.md: C# Implementation Details updated with hosting library pattern - Include manual TOC, index, and overview updates * fix: update next-step links to match new TOC ordering - Agents: running-agents→multimodal→structured-output→background-responses→rag→declarative→observability→tools - Integrations: index→azure-functions→openai-endpoints, m365→a2a - Workflows: branching→agents-in-workflows→human-in-the-loop * Add workflow core concept pages (executors, edges, events, workflows) - Create executors.md, events.md, workflows.md as new concept pages - Rename branching.md to edges.md with conceptual intro and edge type reference - Update TOC.yml with new pages as first entries in Workflows section - Update workflows/index.md Core Concepts to link to new pages - Add redirect for branching.md → edges - Update old core-concepts redirects to point to new pages - Update next-step chain: Overview → Executors → Edges → Events → Workflows → Agents in Workflows * updated language in workflows index * Fix redirect_document_id conflicts and add overview redirect - Set redirect_document_id to false on duplicate entries (42 fixes) - Add redirect for overview/agent-framework-overview.md -> overview/ * Add redirects from user-guide/devui/* to devui/* * Update TOC: add GitHub repo link, move DevUI under Integrations, split Azure Functions/Durable Task * Fix build warnings: invalid link, absolute URLs, alt text, redirect paths * Revert TOC changes, update productDirectory in index.yml instead - Revert TOC.yml to pre-change state (DevUI stays top-level in TOC) - Update index.yml productDirectory: - Add GitHub repo link under Overview - Rename 'Agent tools' -> 'Tools', add 'Providers' under Agents - Add Executors, Edges, Workflow Builder links under Workflows - Add Azure Functions, Durable Task under Integrations - Move DevUI link under Integrations * Update Python custom agent snippet to match current BaseAgent protocol - Single run() method with stream parameter instead of separate run/run_stream - Add proper imports and normalize_messages usage - Show _notify_thread_of_new_messages pattern - Match code repo custom_agent.py sample * Consolidate productDirectory: merge Overview into Get Started, Migration Guide into Support * Set redirect_document_id to false for unpublished page redirects * Update Azure OpenAI and OpenAI provider pages with all three client types Azure OpenAI: - Restructure from ChatCompletion-only to cover Chat, Responses, and Assistants - Add env vars per client type (CHAT_DEPLOYMENT_NAME, RESPONSES_DEPLOYMENT_NAME) - Add AzureOpenAIResponsesClient with hosted tools examples - Add AzureOpenAIAssistantsClient with async-with lifecycle pattern - Add common features section (function tools, threads, streaming) OpenAI: - Split env vars by client type - Fix Assistants to use OpenAIAssistantProvider pattern - Fix streaming to use stream=True parameter * Update Python streaming API: run_stream -> run(stream=True) with ResponseStream - Replace all run_stream() references with run(..., stream=True) - Add ResponseStream documentation with three consumption patterns: async iteration, direct finalization, and combined usage - Update streaming examples in agent run options and response types sections * Update agent.run_stream() to agent.run(stream=True) across docs - Updated agent streaming API in provider pages (anthropic, azure-ai-foundry, github-copilot, custom) - Updated custom.md protocol definition and BaseAgent example to use run(stream=True) pattern - Updated structured-output.md, agents/index.md - Updated ag-ui integration pages (getting-started, backend-tool-rendering, human-in-the-loop, state-management, index) - Updated migration guides (from-semantic-kernel, from-autogen) - Updated workflows/as-agents.md (WorkflowAgent uses agent API) - Left workflow.run_stream() unchanged (valid workflow API) * Fix structured output streaming to use ResponseStream.get_final_response() - Replace AgentResponse.from_update_generator with ResponseStream pattern - Stream's built-in finalizer handles structured output parsing automatically - Show both iterate+finalize and direct finalize patterns - Remove duplicate code block in complete example * Add Python background responses documentation - Replace 'coming soon' placeholder with full Python docs - Document non-streaming polling pattern with continuation_token - Document streaming with resumption pattern - Link to background_responses.py sample * Move semantic-kernel version note after first usage in RAG page * Fix declarative agents: add inline YAML example, remove duplicate snippet * Fix observability: replace removed enable_console_exporters param with env var * Add Claude Agent column to Python tools provider support matrix * Fix GitHub Copilot: enable Local MCP Tools in Python support matrix * Add introductory content and code examples to conversations/index.md * Restructure middleware/index.md Python section - Add intro listing 3 middleware types (matching C# structure) - Reorganize: Function-Based/Class-Based are now subheaders under each type - Move decorators to own section after the three middleware types * Split productDirectory: add Overview header and Get Started with steps 1-6 * Use Learn.microsoft.com content-type SVGs for productDirectory images - Overview: overview.svg - Get Started: getstarted.svg - Agents: concept.svg - Workflows: architecture.svg - Integrations: deploy.svg - Support: howtoguide.svg * Use minihub.svg for Integrations card in productDirectory * Add direct issue and discussion links to support/index.md * removed images that are now linked * Add product directory note and image reference to TOC skill * Replace 'capital' with 'largest city' in agent framework samples --------- Co-authored-by: Dmytro Struk <13853051+dmytrostruk@users.noreply.github.com>
* Update Agent Framework docs for latest Python API changes * Fix AgentSession references in AutoGen migration guide * Align BaseAgent migration example with SupportsAgentRun * Simplify supported providers chat history table * Rename RAG page label and add memory overview link * Refine agents overview tool and streaming sections * Reorganize Conversations & Memory documentation * Add redirects for moved conversation pages * Remove redirected conversation pages * Expand storage guidance from legacy chat-history content * Redirect chat-history URL to storage page * make tip python only * updated redirect
Contributor
Author
|
Learn Build status updates of commit 6b404b1: ✅ Validation status: passedFor more details, please refer to the build report. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
7 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Learn Build Service tried pushing to main, but failed due to branch protection rule in this branch. We create this pull request for manual merge. Please merge this pull request to complete this sync.
If you want to make push successfully without creating such pull requests in future, please follow working with branch protection rule to update your branch protection rule.