Agent Framework Updates & Release Notes

What's Changed

• Python: Add MCPStreamableHTTPTool security guidance for custom http client by @TaoChenOSU in https://github.com/microsoft/agent-framework/pull/7245
• Harden workflow credential selection by @moonbox3 in https://github.com/microsoft/agent-framework/pull/7249
• Python: preserve Gemini 3 thought_signature across function-call replays by @giles17 in https://github.com/microsoft/agent-framework/pull/7095
• .NET: [BREAKING] Hosting OpenAI Responses protocol helpers and optional execution state by @rogerbarreto in https://github.com/microsoft/agent-framework/pull/7000
• .NET: Fix declarative autosend output by @alliscode in https://github.com/microsoft/agent-framework/pull/7217
• .NET: Updating dotnet version for release. by @alliscode in https://github.com/microsoft/agent-framework/pull/7265
• .NET: Added GettingStarted example demonstrating Dapr as an agent provider by @WhitWaldo in https://github.com/microsoft/agent-framework/pull/1615
• Python: Support prompt cache breakpoints for GPT-5.6 models in OpenAI clients by @Mordris in https://github.com/microsoft/agent-framework/pull/7163
• .NET: Fix expensive logging by @alliscode in https://github.com/microsoft/agent-framework/pull/7268
• Python: Fix stateless replay of reasoning-paired tool calls by @moonbox3 in https://github.com/microsoft/agent-framework/pull/7233
• Reduce workflow credential exposure by @moonbox3 in https://github.com/microsoft/agent-framework/pull/7270

New Contributors …

[1.12.0] - 2026-07-21

Added

• agent-framework-azure-cosmos-memory: Add an alpha Azure Cosmos DB semantic-memory context provider with fact extraction, user profiles, samples, and integration coverage (#6719)
• agent-framework-azurefunctions, agent-framework-core, agent-framework-durabletask: Add HITL response-URL addressing for requests raised from inside workflows (#7001)
• agent-framework-core: Add cross-session origin attribution to context-injected messages (#7041)
• agent-framework-core, agent-framework-tools: Warn when auto-approved tools have name collisions (#7090)
• agent-framework-core: Add a session_provider option to MCPSkillsSource and MCPSkill (mutually exclusive with client) that resolves the MCP session on every fetch, keeping cached skills reconnect-safe when the underlying session is replaced (#7135)
• agent-framework-hosting-a2a: Add app-owned A2A hosting helpers (#7050) …

What's Changed

• .NET: Fix CosmosChatHistoryProvider: omit ttl when MessageTtlSeconds is nul… by @TheovanKraay in https://github.com/microsoft/agent-framework/pull/7030
• .NET: Fix CompactionMessageIndex.IsSummaryMessage by @pwoosam in https://github.com/microsoft/agent-framework/pull/7042
• .NET: Fix workflow session bug by @peibekwe in https://github.com/microsoft/agent-framework/pull/7032
• .NET: Enable Valkey NuGet package publishing by @westey-m with @Copilot in https://github.com/microsoft/agent-framework/pull/7059
• .NET: [BREAKING] Graduate message injection out of experimental by @westey-m in https://github.com/microsoft/agent-framework/pull/7044
• .NET: [BREAKING] Graduate todo and agent mode providers out of experimental by @westey-m in https://github.com/microsoft/agent-framework/pull/7052
• Update issue triage by @moonbox3 in https://github.com/microsoft/agent-framework/pull/7098
• .NET: Add name collision warnings for auto-approvals by @westey-m in https://github.com/microsoft/agent-framework/pull/7089
• .NET: Update Microsoft Foundry branding by @nicholasdbrady in https://github.com/microsoft/agent-framework/pull/6999
• .NET: [BREAKING] Harness: Switch FileAccess to opt-in by @westey-m in https://github.com/microsoft/agent-framework/pull/7093
• fix typos in XML doc comments, ADR docs, and test comments by @rinceyuan in https://github.com/microsoft/agent-framework/pull/7085 …

[1.11.0] - 2026-07-09

Added

• agent-framework-core: Add message injection middleware so tools or host code can enqueue messages into an active run and drain them into the next model call within the same AgentSession (#6998)
• agent-framework-core: Integrate message injection into create_harness_agent and the harness console sample so a running harness agent can be nudged mid-turn (#7027)
• agent-framework-core: Add progressive MCP disclosure so agents can discover, load, and unload MCP tool schemas on demand while keeping the allowed_tools boundary intact (#6850)
• agent-framework-core: Add refresh_interval (TTL) to CachingSkillsSource so cached skill lists expire and re-fetch on a configured interval (#6977)
• agent-framework-core, agent-framework-foundry-hosting: Add SkillsSourceContext (invoking agent plus optional session) threaded through the skills source pipeline, enabling context-aware filtering and per-key cache isolation (#6895)
• agent-framework-core: Allow disabling approval for SkillsProvider tools (#6867) …