Back to feed
News Story
量子位
1 sources

openJiuwen Launches Personal Knowledge Space, Natively Supported by WorkSwarm

openJiuwen has launched a Personal Knowledge Space, with WorkSwarm becoming the first to natively support it. The space automatically collects content from sources such as local folders, Edge bookmarks, Zhihu columns, Toutiao, Feishu docs, GitHub, and AtomGit, and organizes it into personal knowledge with a table of contents, connections, and provenance using rule-based, hybrid rule-LLM, or agent-based processing. Agents access this knowledge on demand through progressive disclosure, aiming to enable 'collect once, reuse repeatedly,' improve task accuracy, and reduce token consumption.

SynthePulse Insight · AI deep readingMembers

Collect Once, Reuse Repeatedly: How openJiuwen's Personal Knowledge Space Reshapes Knowledge Supply for Office Agents

Version 1 · 1 source

openJiuwen has launched its Personal Knowledge Space for the first time in WorkSwarm, automatically organizing scattered personal materials into knowledge that can be reused over the long term, and making it available for agents to read on demand through progressive disclosure. A vendor-built 25-question test shows a 32% increase in task pass rate, a 76.2% reduction in total time, and an 84.1% reduction in token consumption during the task phase, but this set of data has not yet been independently verified by a third party.

  • The Personal Knowledge Space is already live in WorkSwarm and can be experienced through a one-click download on the official website, operating around a four-layer flow of sources, collection, processing, and use.
  • Users can add sources such as local folders, Edge bookmarks, Zhihu columns, Toutiao, Feishu documents, GitHub, and AtomGit, and separately configure which materials enter the knowledge space.
  • The system supports manual or scheduled incremental collection, identifying newly added or changed materials by combining source identifiers, versions, and collection progress.
Open section navigationFrom "Starting Over Every Time" to "Collect Once, Reuse Repeatedly"

From "Starting Over Every Time" to "Collect Once, Reuse Repeatedly"

Office agents can already write weekly reports and analyze proposals, but every time a task begins, users still have to search for materials and explain the background, and the agent also has to go through multiple steps such as searching, reading, and organizing before it can truly start work. Switch to another task, and this whole process has to be repeated. Repeatedly processing existing materials wastes time and also increases token consumption; the more materials there are and the longer the steps, the easier it is to miss information or introduce deviations, affecting the accuracy of task completion.

The solution openJiuwen offers is to natively support a "Personal Knowledge Space" for the first time in WorkSwarm. The product is already live in WorkSwarm, and users can experience it through a one-click download on the official website. Its core proposition is "collect once, reuse repeatedly": turning the personal materials users accumulate in daily life into reusable task background knowledge, so that agents can more easily find and make good use of relevant information.

Four-Layer Flow: Sources, Collection, Processing, Use

The Personal Knowledge Space unfolds around four layers: sources, collection, processing, and use. Users first choose personal knowledge sources and collection scope, and the system acquires content on demand or on a schedule; the collected materials undergo format unification and incremental judgment to identify the new content and changes that need to be processed in the current round; then a processing method is selected according to configuration, organizing them into thematic directories and Markdown pages, and maintaining document associations and provenance; after the directory structure and source citation checks pass, the organized results are available for users to view. When the user enables task use, the system provides the personal knowledge overview and reading entry points to the agent, which then gradually reads the relevant pages around the current problem.

The key to this flow is moving "organization" forward out of task execution. Materials are collected and organized in advance, so relevant documents can be consulted directly during the task, reducing the steps of temporarily searching, reading, and organizing raw materials.

Free for now

Read the full analysis

5 more sections of analysis, plus the full takeaway

Loading

Credibility boundary

All facts in this article come from the openJiuwen Personal Knowledge Space launch article published by QbitAI under authorization. Product features, flows, and configuration details are explicitly stated in the article and can be regarded as confirmed product descriptions; the three effect data points of a 32% pass rate increase, 76.2% time reduction, and 84.1% token reduction come from the openJiuwen team's self-built 25-question test set, not an independent third-party evaluation, and the calculation basis, baseline settings, and statistical significance have not been disclosed, so this article labels them as vendor-stated data. The future function of combining user usage trajectories to identify knowledge gaps is only a plan, with no release time or implementation details yet.

Sources for this version

  1. 告别资料散落，让知识持续积累：openJiuwen首发个人知识空间，重塑个人办公智能体新体验！量子位

Primary report

量子位

Primary source