AgentScope Files

Highlight

• Realtime Voice Agent: Introduced a realtime voice agent with DashScope backends and local transport support.
• A2A Protocol Support: Added A2AAgent class to the agent module, enabling agent-to-agent communication.
• New Pipeline Module: Added a new pipeline module with a GoalPipeline class for goal-driven agent workflows.
• Agent-Driven Context Compression: Introduced a context compression tool that allows agents to manage their own context window.
• LLM Rerank for RAG: Added LLM-based reranking support to improve RAG retrieval quality.
• Volcengine Ark Model: Added support for Volcengine Ark chat model.
• MCP Enhancements: Support runtime MCP headers in Docker workspaces and improved MCP connection cleanup.
• Web UI Polish: Chat experience improvements including session auto-naming, query caching, and better dark mode support.

What's Changed

• fix(formatter): derive audio format from media type by @ump45nose in https://github.com/agentscope-ai/agentscope/pull/2301
• fix(gemini): exclude tool-execution tokens from output_tokens by @iluv7 in https://github.com/agentscope-ai/agentscope/pull/2406
• feat(formatter): use native multimodal tool outputs by @fancyboi999 in https://github.com/agentscope-ai/agentscope/pull/2389
• docs(examples): fix broken relative link to _types.py in mem0 README by @feizhuzheng in https://github.com/agentscope-ai/agentscope/pull/2438
• feat(pipeline): add new pipeline module and a GoalPipeline class by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2428
• fix(webui): isolate sessions when switching agents by @iluv7 in https://github.com/agentscope-ai/agentscope/pull/2429
• chore: bump version to 2.0.7.post1 by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2452
• feat(agent): generate a final text summary after reaching max iterations by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/2443
• fix(tracing): report interrupted chat responses by @fengting124 in https://github.com/agentscope-ai/agentscope/pull/2450
• fix(model): preserve encrypted reasoning items in OpenAI Responses history by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/2426
• fix(gemini): sanitize nullable type arrays by @helloJamest in https://github.com/agentscope-ai/agentscope/pull/2437
• fix(webui): keep channel toggle inside card by @iluv7 in https://github.com/agentscope-ai/agentscope/pull/2466
• fix(webui): follow system dark mode so sidebar buttons stay visible (#2467) by @PiedPiper911 in https://github.com/agentscope-ai/agentscope/pull/2468
• fix(state): preserve LRU cache under concurrent hits by @fengting124 in https://github.com/agentscope-ai/agentscope/pull/2457
• fix(xai): include reasoning tokens in output usage by @iluv7 in https://github.com/agentscope-ai/agentscope/pull/2461
• ci(monitor): stop watching review events that can never access secrets by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2472
• fix(formatter): derive multimodal tool-result identifier from the block's own stable id by @AmirF194 in https://github.com/agentscope-ai/agentscope/pull/2165
• fix(example): disable uvicorn hot reload on Windows by @dvd233 in https://github.com/agentscope-ai/agentscope/pull/2434
• fix(tool): make read cache backend-aware so Read->Edit works in workspaces by @nolanchic in https://github.com/agentscope-ai/agentscope/pull/2092
• fix(agent): hint after repeated tool errors by @he-yufeng in https://github.com/agentscope-ai/agentscope/pull/1816
• fix(app): drain the wake-up queue atomically by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2476
• fix(scheduler): let one node own the timers by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2477
• feat(rag): add LLM rerank support by @BreezyB1n in https://github.com/agentscope-ai/agentscope/pull/1975
• fix(scheduler): validate cron before persistence by @redSun64 in https://github.com/agentscope-ai/agentscope/pull/2442
• feat(agent): add agent-driven context compression tool by @jinliyl in https://github.com/agentscope-ai/agentscope/pull/2143
• feat(channel): bind channel credentials interactively by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2484
• fix(channel): show a channel that is starting as connecting, not stopped by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2487
• feat(a2a): add A2AAgent class in the agent module by @YingchaoX in https://github.com/agentscope-ai/agentscope/pull/2142
• docs(middleware): fix reply budget example class name by @zjncs in https://github.com/agentscope-ai/agentscope/pull/2493
• fix(tool): apply schema-guided repair to tool arguments by @iluv7 in https://github.com/agentscope-ai/agentscope/pull/2496
• fix: correct misleading docstring in src/agentscope/init.py by @sunrong1 in https://github.com/agentscope-ai/agentscope/pull/2502
• fix(rag): normalize distance scores so higher means more relevant by @lyh7c00 in https://github.com/agentscope-ai/agentscope/pull/2486
• fix(agent): record compression usage in context by @fengting124 in https://github.com/agentscope-ai/agentscope/pull/2433
• feat(web_ui): chat polish, session auto-naming, and a query cache by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2503
• fix(webui): auto-open task panel on first stateUpdated per session by @aniruddhaadak80 in https://github.com/agentscope-ai/agentscope/pull/2431
• fix(mcp): clean up cancelled MCP connections by @014-code in https://github.com/agentscope-ai/agentscope/pull/2499
• chore(release): bump version to 2.0.8 and add the A2A news item by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2505
• fix: repair agent-facing prompt template text defects by @simpleqt in https://github.com/agentscope-ai/agentscope/pull/2513
• docs: replace full-width colons with ASCII colons in CONTRIBUTING_zh.md URLs by @simpleqt in https://github.com/agentscope-ai/agentscope/pull/2511
• fix(app): clear inbox consumer on session delete by @CoralGarden52 in https://github.com/agentscope-ai/agentscope/pull/2519
• feat(mcp): support runtime MCP headers in Docker workspaces by @qbc2016 in https://github.com/agentscope-ai/agentscope/pull/2456
• fix: block sending by Enter while a reply is in progress by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2529
• fix(rag): escape Excel Markdown table cells by @MaxFreedomPollard in https://github.com/agentscope-ai/agentscope/pull/2528
• fix(deps): pin ripgrep below 15.x to restore Windows dev installs by @Shxiao101 in https://github.com/agentscope-ai/agentscope/pull/2526
• refactor(app): a session's origin is one tagged union by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2536
• fix(app): support cross-owner agent invites by @logicwu0 in https://github.com/agentscope-ai/agentscope/pull/2326
• test(mcp): follow SDK transport defaults by @iluv7 in https://github.com/agentscope-ai/agentscope/pull/2542
• feat(model): add Volcengine Ark chat model support by @iluv7 in https://github.com/agentscope-ai/agentscope/pull/2532
• feat(realtime): realtime voice agent with DashScope backends and local transport by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2547
• docs(zh): link zh news entries to the zh docs locale by @ConcaSun in https://github.com/agentscope-ai/agentscope/pull/2546
• fix(pipeline): honor verifier context reset by @yaodong-shen in https://github.com/agentscope-ai/agentscope/pull/2544
• fix(web-ui): keep running MCP tool calls visible by @yang0228 in https://github.com/agentscope-ai/agentscope/pull/2490
• fix(state): skip oversized read-cache entries by @wuwenbo0626 in https://github.com/agentscope-ai/agentscope/pull/2534
• refactor(realtime): align RealtimeAgent naming with Agent by @DavdGao in https://github.com/agentscope-ai/agentscope/pull/2550
• fix(task): hide completed prerequisites from blockedBy by @wuwenbo0626 in https://github.com/agentscope-ai/agentscope/pull/2541
• fix(rag): preserve blank lines between paragraphs in WordParser by @liugy789 in https://github.com/agentscope-ai/agentscope/pull/2538

New Contributors

• @ump45nose made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2301
• @feizhuzheng made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2438
• @fengting124 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2450
• @PiedPiper911 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2468
• @AmirF194 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2165
• @dvd233 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2434
• @redSun64 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2442
• @zjncs made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2493
• @sunrong1 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2502
• @lyh7c00 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2486
• @aniruddhaadak80 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2431
• @014-code made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2499
• @simpleqt made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2513
• @CoralGarden52 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2519
• @MaxFreedomPollard made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2528
• @Shxiao101 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2526
• @ConcaSun made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2546
• @yaodong-shen made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2544
• @yang0228 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2490
• @wuwenbo0626 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2534
• @liugy789 made their first contribution in https://github.com/agentscope-ai/agentscope/pull/2538

Full Changelog: https://github.com/agentscope-ai/agentscope/compare/v2.0.7...v2.0.8