v1.1.5

1.1.5 (2026-05-19)

Features

• Tools: Added StreamingProgressTool with matching ToolsProcessor plumbing so tools can surface intermediate progress as partial=True events while still emitting a single final function_response; included BaseTool streaming hooks, the llmagent_with_streaming_progress_tool example and verification script.

• Eval: Added RemoteEvalService to drive evaluations against agents exposed over remote interfaces, refactored AgentEvaluator to support remote agent calls, and expanded English/Chinese evaluation docs.

• Model: Landed the OpenAI-compatible adapter layer (models/openai_adapter/{_base,_deepseek,_hunyuan}.py) that isolates provider-specific behavior from OpenAIModel, including DeepSeek v4 thinking / response_format / reasoning_content / token usage handling and hy3-preview ToolPrompt text parsing with streaming filter.

• Examples: Added examples/mempalace_mcp (MemPalace via MCP) and updated examples/llmagent_with_thinking to enable add_tools_to_prompt only for hy3-preview and display thinking / tool calls / final answer separately.

• Utils: Added json_loads_repair and json_repair_string helpers (backed by json_repair) under trpc_agent_sdk.utils, with full unit test coverage.

• Model/Tools: Adopted json_repair only on JSON-tolerant paths — JsonToolPrompt / XmlToolPrompt parse_function, non-streaming OpenAI tool-call args, AgentTool structured-output validation, skills tool result parsing — while keeping strict json.loads for the streaming tool-call accumulator (to preserve "wait for next chunk" semantics) and Hunyuan plain-text <arg_value> parsing (to avoid silently coercing plain text into empty strings).

• Model: Fixed ToolPrompt streaming parsing so multiple tool calls in a single response are all preserved instead of only the last one being kept.

Bug Fixes

• Teams: TeamAgent now honors actions.skip_summarization from custom tool events, so tools like AgentTool(skip_summarization=True) and StreamingProgressTool(skip_summarization=True) end the leader loop without an extra summarization turn (previously masked by leader's disable_react_tool=True).

v1.1.4

Bug Fixes

• Tools: Removed default mempalace_tool exports from trpc_agent_sdk.tools to avoid forcing MemPalace optional dependencies during base package import.

v1.1.3

Features

• Model: Added an OpenAI-compatible adapter layer to isolate provider-specific behaviors, including DeepSeek v4 reasoning/format handling and hy3-preview tool-prompt parsing support.
• Memory: Added MemPalace integration with MemPalaceMemoryService and mempalace_tool, plus related examples and documentation.
• Code Execution: Added Cube/E2B sandbox executor and workspace runtime with optional dependency support and end-to-end example coverage.
• Eval: Added support for evaluating the same metric across different LLMs.

Bug Fixes

• Model: Fixed ToolPrompt streaming parsing so multiple tool calls in one response are preserved instead of only the last call.
• Storage: Improved SQL storage compatibility by filtering empty content parts, fixing MySQL DynamicPickleType serialization, and stabilizing session timestamp updates.
• Eval: Fixed judge-agent JSON output handling in the eval module.
• CI: Added missing e2b-code-interpreter test dependency to prevent cube test collection failures.

1.1.2.post1

Features

• Session: Updated session summarization to retain full conversation history while marking summarized events as model-invisible.
• Session: Added backend-threaded summarization execution to avoid blocking front-end conversation turns.
• Skill: Added multi-user support for skill operations.

Bug Fixes

• Code Execution: Fixed the conflict between code execution and tool invocation where tool data could be lost after code execution.
• MCP: Added support for parsing and returning multiple MCP tool results in a single response.

v1.1.2

Features

• Telemetry: Added OpenTelemetry metrics reporting and introduced custom_metrics to support framework metric reporting when parsing remote agent responses.
• Tools: Added web_search with DuckDuckGo and Google providers, and added web_fetch for webpage content retrieval.
• Docs/Examples: Added usage documentation and examples for web_search and web_fetch.

v1.1.1

Features

• Storage: Added the usage_metadata field to SQL storage and introduced automatic migration for missing columns.
• Skill: Added cross-session skill state persistence so loaded skills can be reused across sessions, reducing repeated skill loading and unnecessary retry turns.
• Skill: Added skill install/uninstall awareness so the model can detect skill lifecycle changes and avoid missing-skill lookups or calls to uninstalled skills.
• Session: Added usage_metadata support in SQL session storage for persisting and reading token usage statistics.

Bug Fixes

• Skill: Reduced hallucinated skill command generation when users intend to run commands, lowering invalid command attempts and retry loops.

trpc-agent-python framework

Features

• Unified Agent framework with LlmAgent, LangGraphAgent and TransferAgent
• Multi-agent orchestration with built-in Chain, Parallel, and Cycle patterns, plus Team and nested Team collaboration
• Human-in-the-loop workflows with pause, review, and resume support for long-running tasks
• Rich tool ecosystem including built-in file/shell tools, MCP tools, LangChain tools, and extensible third-party integrations
• Extensible Skill system with local and HTTP distribution, dynamic loading, timeout control, and sandbox execution
• Code execution support with async runtime and sandbox/container execution options
• Session and memory services with in-memory, Redis, and SQL backends, including filtering, summarization, and scheduled cleanup
• RAG and knowledge capabilities through LangchainKnowledge with loaders, splitters, embedders, vector stores, retrievers, and prompt templates
• Evaluation framework with trajectory and response quality assessment, LLM-judge metrics, parallel evaluation, and JSON reporting
• Service and protocol integrations for A2A, AG-UI, and OpenClaw runtime scenarios
• OpenClaw runtime capabilities for gateway/chat/ui/deps workflows with pluggable channels, tools, skills, session and memory integration
• OpenClaw skill dependency management with profile-based inspection and install planning for common runtime environments
• Observability via tracing support, including end-to-end execution flow, tool-call traces, and cancellation traces
• Developer experience support with practical examples and DebugServer for local development and validation