# multi-agent orchestration

Recursa splits adaptation to a new domain into two steps: it bootstraps a navigable knowledge repository from domain materials, then has the same model act as both solver and learner, with parallel child learners diagnosing, editing and validating revisions to instructions, skills and memory. On Harvey LAB's Firm Knowledge legal benchmark, all-pass rate on held-out test tasks rises from 30.0% to 55.0% with DeepSeek-V4.1-Flash as solver and from 35.0% to 46.0% with Qwen3.8-Flash, with no stronger model or human expert writing the revisions.

Recursa splits adaptation to a new domain into two steps: it bootstraps a navigable knowledge repository from domain materials, then has the same model act as both solver and learner, with parallel child learners diagnosing, editing and validating revisions to instructions, skills and memory. On Harvey LAB's Firm Knowledge legal benchmark, all-pass rate on held-out test tasks rises from 30.0% to 55.0% with DeepSeek-V4.1-Flash as solver and from 35.0% to 46.0% with Qwen3.8-Flash, with no stronger model or human expert writing the revisions.

OpenMAIC is an open-source multi-agent platform that turns any topic or document into an interactive classroom, with AI teachers and AI classmates who speak, draw on a whiteboard and discuss in real time. Version 1.3.0 keeps each upload in its owner's knowledge base, organized in folders and parsed once for every conversation, where the agent can read, search and organize it and the composer offers it through @. The Pro workbench also previews the agent's work before the first page exists, and classroom tabs switch between Learn and Edit.

OpenMAIC is an open-source multi-agent platform that turns any topic or document into an interactive classroom, with AI teachers and AI classmates who speak, draw on a whiteboard and discuss in real time. Version 1.3.0 keeps each upload in its owner's knowledge base, organized in folders and parsed once for every conversation, where the agent can read, search and organize it and the composer offers it through @. The Pro workbench also previews the agent's work before the first page exists, and classroom tabs switch between Learn and Edit.

DataMagic introduces the declarative specification DVSpec and a "Generate-then-Orchestrate" multi-agent strategy that authors multi-scene data videos with dynamic charts, voice narration, and synchronized animation from raw tabular data, raising video quality from 2.13/5 for the strongest direct LLM generation to 3.89/5 and execution success from 48.62%-86.24% to above 95% on 109 real-world samples, while a 12-participant user study shows about a 79.7% reduction in task time.

DataMagic introduces the declarative specification DVSpec and a "Generate-then-Orchestrate" multi-agent strategy that authors multi-scene data videos with dynamic charts, voice narration, and synchronized animation from raw tabular data, raising video quality from 2.13/5 for the strongest direct LLM generation to 3.89/5 and execution success from 48.62%-86.24% to above 95% on 109 real-world samples, while a 12-participant user study shows about a 79.7% reduction in task time.

Raven is an open-source multi-agent ecosystem that treats each executable model–harness pair as a composable unit of intelligence: a Host Agent decomposes goals into a directed acyclic graph and matches subtasks to registered specialists, HarnessBank-style evolution, EverOS memory, and Skill Forge reuse experience across tasks, and the paper gives sufficient conditions under which composition expands reliable task coverage under a shared resource budget, reporting gains over Claude Code, Hermes Agent, OpenClaw and other baselines on the MAOB planning benchmark, four native specialist domains, frozen-backbone harness evolution, and fixed skill-library reuse.

Raven is an open-source multi-agent ecosystem that treats each executable model–harness pair as a composable unit of intelligence: a Host Agent decomposes goals into a directed acyclic graph and matches subtasks to registered specialists, HarnessBank-style evolution, EverOS memory, and Skill Forge reuse experience across tasks, and the paper gives sufficient conditions under which composition expands reliable task coverage under a shared resource budget, reporting gains over Claude Code, Hermes Agent, OpenClaw and other baselines on the MAOB planning benchmark, four native specialist domains, frozen-backbone harness evolution, and fixed skill-library reuse.

NVIDIA announced the Open Agent Safety Platform, an open software platform and reference system design, and collaborated with Anthropic to combine Claude Managed Agents with the open source NVIDIA OpenShell runtime: Managed Agents keeps the passwords and access keys an agent needs in a separate vault so the agent never sees them and adds audit trails plus integration with existing access controls, while OpenShell enforces policies outside the agent for every tool, file, network connection and data access, blocking everything unless a rule allows it and logging every decision it allows or blocks, so teams can start with narrow permissions, review the log, tighten rules toward least access with Claude, and use the policy prover to confirm by mathematical proof what the agent can reach under

NVIDIA announced the Open Agent Safety Platform, an open software platform and reference system design, and collaborated with Anthropic to combine Claude Managed Agents with the open source NVIDIA OpenShell runtime: Managed Agents keeps the passwords and access keys an agent needs in a separate vault so the agent never sees them and adds audit trails plus integration with existing access controls, while OpenShell enforces policies outside the agent for every tool, file, network connection and data access, blocking everything unless a rule allows it and logging every decision it allows or blocks, so teams can start with narrow permissions, review the log, tighten rules toward least access with Claude, and use the policy prover to confirm by mathematical proof what the agent can reach under