A RESEARCH TOOLKIT FOR AI AGENTS
Help AI agents write research reports with concrete judgments, traceable sources, and independent review.
Creating a file, editing an existing object, and changing business state are different deliverables. A report is checked mainly for content quality; updating a sales ledger must preserve fields and calculations; sending a quotation or submitting an approval changes other people's work.
Report excerpt
Changes in one office-agent report: clearer prose, added facts, and narrower inferences.
The opening of the workspace section moves from a broad value claim to how the products organize work.
The final report shortens the rhetorical setup and a vague closing, then names tools, context, and device availability.
The value of independent office agents lies not in moving a chat box onto the desktop, but in bringing materials, execution, and revisions into one stretch of work. For people who frequently organize files from different sources, this approach can reduce uploading, downloading, and switching tools. Yet keeping files on a computer does not mean tasks run continuously or all processing happens offline. The differences between Kimi Work and Coze show that independent workspaces do not share a single implementation.
The opening frames value through a rhetorical contrast; the closing only says implementations differ.
Original passage (Chinese) ↗Independent office agents organize materials, execution, and revisions in one workspace. For people who frequently organize files from different sources, this approach can reduce uploading, downloading, and switching tools. Continuous operation and the location of data processing still depend on the implementation. The differences between Kimi Work and Coze center on who supplies tools, who retains context, and who keeps the device online.
The report states the mechanism directly and identifies comparison criteria, while retaining its conditions.
Original passage (Chinese) ↗Historical report. Information through September 9, 2026; some dynamic documents were checked on September 10. Public-source research without installed-product testing. These versions are not a controlled with/without-toolkit experiment.
RESEARCH REQUEST
Send this request to an agent that can read the repository and access your sources.
If repository access is unavailable, upload the files or paste the instructions. Native Skill installation is optional.
Ways to provide the toolkitUse Research Toolkit for this task: https://github.com/rrrrrredy/research-toolkit Research office agents available to users in China in 2026. Compare Kimi Work, Coze, Feishu / Doubao Work, Wukong, WPS, and WorkBuddy. Write for AI product and workplace leads. Explain which files and business objects each product changes, where tasks run, and how models, tools, context, permissions, and human involvement affect delivery and rework. Use information available as of [DATE]. Cite important claims. Distinguish company statements, media experiences, and measured outcomes. Explain what could overturn your conclusions. Read SKILL.md. Clarify missing requirements and agree on an outline before collecting sources.
Replace [DATE] and adapt the topic and audience. This request is based on the archived case brief; a new run may reach different conclusions.
File and text guideFrom the brief to the report, each stage has a concrete output.
Audience, scope, decisions, and an agreed outline.
Research brief ↗Primary sources, access limits, and contrary evidence.
Source records ↗Link facts to explanations, tradeoffs, and implications.
Claim records ↗Examine the whole report and its supporting evidence.
Review records ↗Concrete judgments, traceable sources, and honest limits.
Complete report ↗The links provide this case's research materials. The historical report did not run through the current MCP version.
Scope, sources, reasoning, writing, and review. Detailed guidance is loaded by your agent when needed.
Create tasks, retain progress, run independent reviews, and check delivery conditions.
Choose the input your interface supports. Plugin and MCP tools are optional.
| Repository link | Provide the repository URL and research request. The agent needs to read SKILL.md and the required references. | Repository use |
|---|---|---|
| Files or attachments | Provide SKILL.md and the reference files from the downloaded repository. Supply further methods when needed. | Download files |
| Pasted text | Paste the full entry instructions and the reference sections needed for the task. A URL alone is insufficient without repository access. | Chat-only use |
| Native Skill | For repeated use, install the complete toolkit through the host’s supported Skill mechanism. | Skill setup |
| Plugin | A bundle of the methods and local MCP tools for hosts that support its package format and local execution. | Installation guide |
|---|---|---|
| MCP | Five local tools for task records, progress, guidance, review calls, and delivery checks. | Connect MCP |
Reading or uploading the methods does not start MCP or run an independent review. The tools need local stdio MCP support; the plugin also requires its package format.
Independent review needs a separate reviewer with the report and evidence. The automated executor defaults to Codex CLI or a configured review command; the author agent does not need to be Codex. Review calls use the configured model account and send it the supplied material.
A recording of the case browser and request copying.