v2.5.10

Changelog

New Features:

• Parallel Search Tool: Add Parallel AI Search support for Vertex AI with native ToolParallelAiSearch integration
• Observability: Extended support with MLflow to have full traces observability of your agents. See docs.
• Docling Reader: Integrated Docling library as a new reader, enabling document processing across a multiple range of file formats.
• WhatsApp Interface V2:
  • Added media support (images, video, audio, documents), interactive messages (reply buttons, list menus, locations, reactions), and Team/Workflow support.
  • Users can send /new to start fresh conversations.
  • Added enable_encryption parameter for encrypting phone numbers
• Telegram Interface:
  • Added new Telegram interfaces for AgentOS with support for agents, teams, and workflows.
  • Multi-modal support and use /new to start fresh conversations.
• Telegram Tools: Added new tools so agents can send photos, documents, videos, audio, animations, stickers among other things

Improvements:

• Workflow:
  • Added version query param to GET /workflows/{id}to fetch specific workflow versions.
  • Added run level params like- metadata, dependencies, add_dependencies_to_context, add_session_state_to_context similar to agent/team.
• Mistral: Added mistralai v2 support while maintaining backward compatibility with v1.
• Gemini: Added timeout parameter to Gemini model class.

Bug Fixes:

• MCPTools: Fixed race condition where parallel MCP tool calls with header_provider would each create duplicate sessions, causing the agent to get stuck.
• Claude: Fixed structured output detection for supported models
• Streaming Tool Calls: Fixed duplicate tool execution in streaming mode caused by shared dict references in parse_tool_calls
• MongoDb: Fixed the incorrect import of the pymongo async modules.

What's Changed

• fix: prevent deadlock in parallel MCP tool calls with header_provider (fixes #6094) by @giulio-leone in #6821
• fix: remove unnecessary fast-path stale session cleanup in MCPTools by @ysolanky in #6949
• feat: add Parallel web search grounding for Vertex AI Gemini by @ysolanky in #6060
• fix: correct async MongoDB import by @v-vinson in #6399
• fix: mongo validation by @kausmeows in #6957
• feat: add version query param to GET /workflows/{id} endpoint by @harshsinha03 in #6956
• fix: async PostgresDb returns None for missing tables instead of logging ERROR by @kausmeows in #6965
• [fix] replace fragile blocklist with prefix-based structured output detection by @hztBUAA in #6643
• cookbook: add MLflow observability integration via OpenInference by @ysolanky in #6930
• chore: add PR triage automation and improve contribution guidelines by @ysolanky in #6959
• fix: support mistral v2 imports by @harshsinha03 in #6969
• feat: add run-level params to Workflow.run() and arun() by @uzaxirr in #6927
• fix: separate chat and responses metrics by @ysolanky in #6973
• feat: add Telegram interface and extended TelegramTools by @Mustafa-Esoofally in #6377
• chore: consolidate pr-triage workflow into single comment by @ysolanky in #7013
• feat: add timeout parameter to Gemini model class by @fehmisener in #7021
• fix: use list comprehension in parse_tool_calls to prevent shared dict references by @ArivunidhiA in #6579
• feat: add Docling reader integration for advanced document processing by @sannya-singal in #6981
• fix: register Telegram interfaces in system test gateway by @Mustafa-Esoofally in #7026
• feat: WhatsApp interface V2 — media I/O, interactive tools, Team/Workflow support by @Mustafa-Esoofally in #6466
• fix: use valid Telegram bot token format in system tests by @Mustafa-Esoofally in #7027
• chore: Release v2.5.10 by @kausmeows in #7011

New Contributors

• @v-vinson made their first contribution in #6399
• @sannya-singal made their first contribution in #6981

Full Changelog: v2.5.9...v2.5.10

v2.5.9

Changelog

New Features:

• Followups: Added built-in followup suggestions for Agent and Team. See cookbook.
• Datetime Format: Added datetime_format parameter to Agent and Team, allowing custom strftime formatting for datetime context (e.g., ISO-8601, date-only, localized).
• Tool Hook Message History: Tool pre/post hooks and agent-level tool_hooks can now access the current run's message history via run_context.messages
• GoogleCalendarTools: Extended GoogleCalendarTools with new tools, service account auth, and cookbooks.

Bug Fixes:

• HITL: Fixed human in the loop to support add_history_to_context during multi-round conversations while running
• Tools: Removed (None) prefix from tool parameter descriptions
• Siliconflow Model Provider: Updated to use correct default base_url from .com to .cn
• Learning PROPOSE Mode: Auto-enable chat history for LearningMode.PROPOSE for multi-turn confirmation.
• AgentOS API Docs: Added descriptions to all Form/File parameters across agent, team, and workflow API endpoints for OpenAPI/Swagger visibility

What's Changed

• fix: resolved continue_run, missing history messages issue by @abhi10691 in #6888
• fix: remove (None) prefix from tool parameter descriptions by @fehmisener in #6906
• feat: add datetime_format to Agent and Team for custom datetime context by @hztBUAA in #6649
• feat: expose run message history to tool hooks with mutation safety by @hztBUAA in #6652
• feat: built-in followup suggestions for Agent and Team by @ysolanky in #6672
• [fix] siliconflow: correct default base_url from .com to .cn by @NIK-TIGER-BILL in #6933
• feat: extend GoogleCalendarTools with new tools, service account auth, and cookbooks by @Mustafa-Esoofally in #6926
• fix: auto-enable chat history for LearningMode.PROPOSE by @Mustafa-Esoofally in #6928
• chore: add form parameter descriptions to AgentOS API endpoints by @uzaxirr in #6941
• chore: Release v2.5.9 by @kausmeows in #6940

New Contributors

• @abhi10691 made their first contribution in #6888
• @NIK-TIGER-BILL made their first contribution in #6933

Full Changelog: v2.5.8...v2.5.9

v2.5.8

Changelog

New Features

• Human-Readable IDs: Agents and teams now use human-readable identifiers (e.g., brave-falcon-7x3k) instead of UUIDs, making debugging and monitoring more intuitive.
• AgentOS: serve() now supports AGENT_OS_HOST and AGENT_OS_PORT environment variables as fallbacks, simplifying container deployments.
• GitlabTools: Added GitlabTools with read-focused GitLab integrations, async support, and cleaner tool configuration.
• GmailTools: Extended GmailTools with new tools and service account auth.

Bug Fixes

• Media in Run Output: Generated media (images, audio) is now always included in run output, regardless of the store_media setting and is rather scrubbed before storing in DB
• OpenAIResponses: Fixed handling of mixed external_execution and regular tools when using OpenAI Responses API.
• MySQL: Added json_serializer to MySQL engine creation for proper JSON handling.
• PostgreSQL: Removed session_type filter from get_session to prevent session data loss when session types change.
• Workflow: Fixed a bug where Loop iterations always received the original input instead of the previous iteration's output. Added a flag forward_iteration_output to opt-in to this behaviour.

What's Changed

• feat: Docker-style human-readable IDs for agents and teams by @ysolanky in #6869
• feat: add AGENT_OS_HOST/AGENT_OS_PORT env var fallbacks to serve() by @Mustafa-Esoofally in #6857
• fix: remove session_type filter from PostgreSQL get_session to prevent session data loss by @kausmeows in #6873
• fix: add json_serializer to MySQL engine creation by @willemcdejongh in #6848
• [fix] handle mixed external_execution and regular tools for OpenAI by @hztBUAA in #6630
• fix: include generated media in run output regardless of store_media setting by @Br1an67 in #6793
• cookbook: add x402scan MCP tools example by @SamOpenClaw in #6555
• feat: add GitlabTools with tests and cookbook example by @SalimELMARDI in #6721
• feat: extend GmailTools with new tools, service account auth, and agentic cookbooks by @Mustafa-Esoofally in #6765
• fix: loop iterations to forward output from previous iteration by @kausmeows in #6868
• fix: resolve duplicate operationId warnings in disabled feature router by @kausmeows in #6887
• chore: Release v2.5.8 by @kausmeows in #6886

New Contributors

• @SamOpenClaw made their first contribution in #6555
• @SalimELMARDI made their first contribution in #6721

Full Changelog: v2.5.7...v2.5.8

v2.5.7

New Features:

• OpenAI-Like Embedder: Added OpenAILikeEmbedder for providers with OpenAI compatible embedding endpoints

Improvements:

• Session Search Tool: Agents and teams can now browse previous sessions with a two-step search_past_sessions + read_past_session pattern. Session previews show per-run user/assistant pairs instead of a single message, with configurable limits on sessions searched (num_past_sessions_to_search) and preview depth (num_past_session_runs_in_search). read_past_session accepts an optional num_runs parameter so the model can avoid pulling in full conversations from long sessions.

Bug Fixes:

• ChromaDB: Added ID deduplication for documents with identical content across all insert and upsert methods.
• Azure OpenAI Embedder: Added support for non-default Azure-deployed models (e.g. Cohere) and improved dimensions handling for custom deployments.
• Slack Interface: Fixed a bug preventing user_id from being propagated

What's Changed

• cookbook: restructure knowledge cookbooks into story-driven journeys by @willemcdejongh in #6681
• [feat] add OpenAILikeEmbedder for LiteLLM proxy compatibility by @hztBUAA in #6648
• fix: wrap input as list for non-default Azure models by @giulio-leone in #6809
• fix: deduplicate ChromaDB document IDs to prevent batch insert failures by @harshsinha03 in #6714
• fix: pass user_id for all Slack events and add thread_message_limit by @Mustafa-Esoofally in #6766
• refactor: simplify search_session_history — drop query, add per-run limits by @ysolanky in #6858
• feat: propagate team reference to member agent tools by @Mustafa-Esoofally in #6707
• chore: release 2.5.7 by @ysolanky in #6864

New Contributors

• @giulio-leone made their first contribution in #6809

Full Changelog: v2.5.6...v2.5.7

v2.5.6

Changelog

New Features:

• GitHub App Authentication for Knowledge Sources: Adds support for authenticating with GitHub as a GitHub App (app_id, installation_id, private_key) in addition to personal access tokens, with thread-safe token caching and both sync/async variants.
• File Upload: Added image/heic and image/heif support to file upload endpoints.
• Approvals: Added approval status endpoint and admin-gated continue run enforcement.
• Traces: Added advanced filtering DSL support in Agent OS

Improvements:

• Google Tools: Restructured Google tools into agno.tools.google sub-package (from agno.tools.google import GmailTools). Old import paths still work with a backwards compatibility.
• Team Task Mode:
  • Enhances TeamMode.tasks streaming to provide structured task data for frontend rendering in events like TeamTaskCreated and TeamTaskUpdated
  • Structured task data in TaskStateUpdatedEvent: Added tasks: List[TaskData] field containing full task details (id, title, description, status, assignee, dependencies, result) and completion_summary field

Bug Fixes:

• OpenAI Responses:
  • Fixed vector store polling race condition where file_search silently returned empty results due to eventual consistency in OpenAI's vector_stores.files.list
    API.
  • Added input_file support to OpenAIResponses
• Sentence Transformer Reranker: Fixed VRAM memory leak in SentenceTransformerReranker caused by CrossEncoder being instantiated on every _rerank() call instead of being reused.
• JSON Schema: Extended support for Literal types in JSON schema generation

What's Changed

• feat: extend Registry with agents/teams for UI workflow rehydration by @ysolanky in #6278
• cookbook: add Gemini 3 progressive guide by @ysolanky in #6730
• fix: add support for Literal types in JSON schema generation by @kausmeows in #6723
• fix: updated links in README by @kyleaton in #6749
• feat: add image/heic and image/heif support to file upload endpoints by @Br1an67 in #6791
• feat: add input_file support to OpenAIResponses by @uzaxirr in #6806
• [fix] properly release SentenceTransformerReranker models to prevent VRAM leak by @hztBUAA in #6638
• feat: add advanced filtering dsl by @kausmeows in #6671
• feat: restructure Google tools, add FileTools.search_content, add enable flags by @ashpreetbedi in #6833
• feat: add GitHub App authentication support to GitHubConfig by @willemcdejongh in #6831
• fix: use Coroutine return types for arun/acontinue_run overloads by @ysolanky in #6836
• feat: add approval status endpoint and admin-gated continue run enforcement by @kausmeows in #6598
• chore: update team tasks stream by @kausmeows in #6718
• fix: return 503 instead of 404 for disabled AgentOS features by @ysolanky in #6837
• chore: Release v2.5.6 by @kausmeows in #6823

New Contributors

• @Br1an67 made their first contribution in #6791

Full Changelog: v2.5.5...v2.5.6

v2.5.5

Changelog

New Features:

• Slack Interface:
  • Responses now stream in real-time with live progress cards for tool calls, reasoning, and workflow steps.
  • Each Slack instance supports its own token and signing_secret, enabling multiple bots on the same server.

Improvements:

• ModelsLabTools: Extended ModelsLabTools to support image generation via ModelsLab's text-to-image API, completing the full ModelsLab media suite.

Bug Fixes:

• Gemini: Fixed empty message parts causing request failures when sending conversations to the Gemini API.
• AWS Bedrock: Merge consecutive toolResult blocks into single user message.
• Workflows: Handle raw image bytes in workflow step’s _convert_image_artifacts_to_images
• Knowledge Filters: serialize FilterExpr objects in GET /agents and /teams response

What's Changed

• [fix] Merge consecutive toolResult blocks into single user message by @Yoo-Joo in #6694
• feat: Add image generation (PNG/JPG) to ModelsLabTools by @adhikjoshi in #6670
• chore: fix links and update to new cookbook structure by @shaloo in #6616
• feat: Slack interface updates by @Mustafa-Esoofally in #6615
• [fix] Handle raw image bytes in workflow step _convert_image_artifacts_to_images by @Himanshu040604 in #6710
• fix: missing registry apis, openapi title and dynamic version pickup by @shaloo in #6713
• [fix] filter empty message parts before sending to Gemini API by @hztBUAA in #6644
• fix: serialize FilterExpr objects in GET /agents response by @uzaxirr in #6581
• feat: add approval and teams agent os cookbook by @Ayush0054 in #6717
• fix: add aiohttp to gateway Dockerfile for Slack async client by @Mustafa-Esoofally in #6719
• chore: Release v2.5.5 by @kausmeows in #6712

New Contributors

• @adhikjoshi made their first contribution in #6670
• @shaloo made their first contribution in #6616
• @Himanshu040604 made their first contribution in #6710

Full Changelog: v2.5.4...v2.5.5

v2.5.4

Changelog

New Features:

• PgVector: added similarity_threshold parameter for filtering search results by minimum similarity score.
• Team: Added streaming event support for TeamMode.tasks, enabling real-time event emission during autonomous task execution.
• Workflows HITL: Add support for HITL in workflows at Step level to pause for confirmation and user input. See docs

Improvements:

• Metrics: Redesigns the metrics system to provide per-model, per-component granular tracking across the entire agent/team/workflow lifecycle.
• WebsearchTools: expose timelimit, region and backend params in DuckDuckGoTools
• SeltzTools: Updated to use Seltz SDK 0.1.3

Bug Fixes:

• AWS Bedrock: Fixed to refresh AWS credentials on every Bedrock request when using boto3 session.
• @tool decorator: Fixed @tool decorator to not fail for Agent/Team type annotations
• Pydantic model serialization: Fixed Pydantic model serialization in tool cache.

What's Changed

• chore: update quickstart by @ashpreetbedi in #6656
• chore: new cookbook by @ashpreetbedi in #6663
• feat: add similarity_threshold parameter for filtering search for PgVector [SDK-441] by @harshsinha03 in #6216
• cookbook: agent level by @ysolanky in #6665
• chore: split team tests into 4 balanced parallel jobs by @ysolanky in #6628
• fix: handle Pydantic model serialization in tool cache by @fehmisener in #6678
• [feat] expose timelimit, region and backend params in DuckDuckGoTools by @hztBUAA in #6650
• fix: @tool decorator fails with Agent/Team type annotations by @ysolanky in #6610
• chore: openai pdf input bytes by @kausmeows in #6685
• fix: add CI guard for split test coverage + fix 3 missing knowledge tests by @Mustafa-Esoofally in #6668
• fix: use fresh OpenAI clients in memori multi-registration tests by @Mustafa-Esoofally in #6691
• fix: refresh AWS credentials on every Bedrock request when using boto3 session by @Mustafa-Esoofally in #6657
• feat: metrics update by @harshsinha03 in #6405
• feat: support streaming for team task mode by @kausmeows in #6622
• [feat] Update SeltzTools for Seltz SDK 0.1.3 by @WilliamEspegren in #6292
• fix: include member response metrics in team session metrics by @ysolanky in #6704
• feat: step level workflow HITL by @kausmeows in #6211
• refactor: replace heavyweight RunOutput collectors with RunMetrics by @ysolanky in #6703
• fix: harden hook execution and PII guardrail by @Mustafa-Esoofally in #6483
• chore: Release v2.5.4 by @kausmeows in #6661

Full Changelog: v2.5.3...v2.5.4

v2.5.3

Changelog

New Features:

• Remote Knowledge: Added endpoints to support listing of remote contents via API. Enables uploading of content in S3 buckets via AgentOS

Improvements:

• PDF Reader: Added sanitize_content parameter to BasePDFReader (enabled by default) that normalizes fragmented PDF text extraction — collapsing word-per-line artifacts while preserving paragraph breaks. Disable with sanitize_content=False for PDFs containing structured content like code or tables.
• PDF Reader: Added a sanitize_content parameter to BasePDFReader that normalizes extracted PDF text by collapsing excessive whitespace sequences into single spaces.

Bug Fixes:

• Knowledge Retriever: Fix issue whereknowledge_retriever ignored when search_knowledge=True enabled.
• Claude: Fixed streaming metrics double-count input_tokens for Anthropic models
• Knowledge Search: Fixed match include/exclude patterns against filename for directory scanning
• Reasoning: fixed reasoning_content duplication across reasoning steps.
• Deepcopy on Agent/Team/Workflow: fixed deep_copy fails for nested teams due to non-init dataclass fields.
• HITL: fixed HITL pause handlers silently dropping session_state changes in both agent and team modules.
• Workflow Parallel Steps: fixed race condition on run_context.output_schema in Parallel execution.

What's Changed

• [fix] Prevent double-counting input_tokens for Anthropic streaming metrics by @ArivunidhiA in #6580
• feat: add remote content sources with cloud storage loading and browsing by @willemcdejongh in #6584
• chore: anchor gitignore paths for symlinked dirs and add /apps by @ashpreetbedi in #6588
• fix: match include/exclude patterns against filename for directory scanning by @themavik in #6455
• fix: prevent reasoning_content duplication across reasoning steps by @themavik in #6438
• fix: deep_copy fails for nested teams due to non-init dataclass fields by @rodrigocoliveira in #6297
• feat: add content sanitization to PDF reader by @willemcdejongh in #6591
• ci: smart @claude handler with review plugin routing by @Mustafa-Esoofally in #6603
• fix: handle missing mime_type in Gemini file formatting by @Mustafa-Esoofally in #6367
• fix: remove prompt input from claude review to enable tag mode by @Mustafa-Esoofally in #6604
• fix: move concurrency to job level to prevent self-cancellation by @Mustafa-Esoofally in #6605
• ci: enable show_full_output for claude review debugging by @Mustafa-Esoofally in #6606
• cookbook: v2.5 testing audit fixes and test logs by @Mustafa-Esoofally in #6546
• [fix] HITL pause session state loss by @Mustafa-Esoofally in #6495
• fix: move Parallel.steps to a proper dataclass field with default_factory by @ysolanky in #6608
• fix: prevent race condition on run_context.output_schema in Parallel execution by @ysolanky in #6609
• fix: knowledge_retriever ignored when search_knowledge=True by @willemcdejongh in #6593
• chore: update yfinance tools to have options by @kausmeows in #6607
• feat: add OpenTelemetry and Agno instrumentation deps to os extras by @ysolanky in #6618
• fix: auto-inject linked_to filter for list-based FilterExpr filters by @willemcdejongh in #6611
• fix: add files field to WorkflowRunOutput and fix file serialization by @kausmeows in #6544
• fix: datetime JSON serialization for PostgreSQL session storage by @kausmeows in #6363
• fix: complete user_id handling in SDK client session methods by @Mustafa-Esoofally in #6502
• feat: add is_component, current_version, and stage to list endpoints by @ysolanky in #6266
• chore: release 2.5.3 by @ysolanky in #6619

New Contributors

• @ArivunidhiA made their first contribution in #6580
• @themavik made their first contribution in #6455

Full Changelog: v2.5.2...v2.5.3

v2.5.2

Changelog

• License Change: Changed license from Mozilla Public License 2.0 (MPL 2.0) to Apache Software License.

What's Changed

• Release 2.5.2 by @ashpreetbedi in #6578

Full Changelog: v2.5.1...v2.5.2

v2.5.1

Changelog

New Features

• CodingTools: New toolkit for code-related operations.
• UserFeedbackTools: New toolkit for collecting user feedback.

Bug Fixes

• Team Member Resolution: Prioritize explicit id over name in get_member_id for more predictable member selection.
• TavilyTools: Removed deprecated include_answer kwarg from search method.
• MCP Transport: Fail explicitly when header_provider is incompatible with transport type.
• FastEmbed: Cache FastEmbed model and reset reader state between calls for better performance.

What's Changed

• [fix] remove include_answer kwarg from tavily tool method by @mklrlnd424 in #5755
• refactor: Fail explicitly when header_provider is incompatible with transport by @bhatt-neel-dev in #5988
• fix: prioritize explicit id over name in get_member_id by @kausmeows in #6336
• fix: Cache FastEmbed model and reset reader state between calls by @Mustafa-Esoofally in #6029
• test: improve test suite with rate-limit protection by @ashpreetbedi in #6535
• cookbook: Restructure 02_agents into numbered directories with feature coverage by @ashpreetbedi in #6536
• cookbook: Improve cookbook demo by @ashpreetbedi in #6538
• cookbook: Make 03_teams fully testable and feature-complete by @ashpreetbedi in #6557
• ci: add Claude Code workflow with code-review plugin by @Mustafa-Esoofally in #6560
• fix: use correct marketplace for code-review plugin by @Mustafa-Esoofally in #6562
• feat: add CodingTools, UserFeedbackTools, and demo cookbook updates by @ashpreetbedi in #6572
• release: v2.5.1 by @ashpreetbedi in #6576

New Contributors

• @mklrlnd424 made their first contribution in #5755

Full Changelog: v2.5.0...v2.5.1