11 role-specific bundles that turn Claude into a specialist for your team

If every role uses different tools yet you re-explain your setup to Claude each time, this is for you. knowledge-work-plugins open-sources 11 job functions — sales, fi...

Quick answer

• Claude is useful when the reader needs the decision frame before the full tutorial.
• The practical answer is: Explain what Claude changes, when it is useful, and how to verify it safely.
• Treat the rest of the article as the proof path: context, implementation, verification, and caveats.

When repeating yourself becomes the job

Sales, customer support, finance, data, marketing. Even inside one company, every role uses different tools and works in different ways. So every time you hand Claude a task, you re-explain it: which tools your team uses, what order you do things in, what a term means in your world. Once or twice is fine. Done daily, that explanation becomes the work itself.

Anthropic's knowledge-work-plugins starts from cutting that repetition. It's a collection of plugins that bundle, ahead of time, what Claude needs to know for a given job function.

The problem it solves

Cowork lets you set a goal and have Claude deliver finished work. The catch is that the definition of 'finished work' differs by role and by team. The call prep a salesperson wants and the close process a finance team needs are entirely different jobs.

knowledge-work-plugins closes that gap with per-role bundles. Each plugin packages the skills, connectors, slash commands, and sub-agents for one specific job function. Install it and you get a strong starting point for anyone in that role.

Eleven plugins are open-sourced: productivity, sales, customer-support, product-management, marketing, legal, finance, data, enterprise-search, bio-research, and cowork-plugin-management for building and customizing plugins themselves.

How it works

Every plugin follows the same structure.

plugin-name/
├── .claude-plugin/plugin.json   # Manifest
├── .mcp.json                    # Tool connections
├── commands/                    # Slash commands you invoke explicitly
└── skills/                      # Domain knowledge Claude draws on automatically

• Skills encode the domain expertise and step-by-step workflows. Claude draws on them automatically when relevant.
• Commands are explicit actions you trigger, like /finance:reconciliation or /product-management:write-spec.
• Connectors wire Claude to the external tools the role depends on — CRMs, project trackers, data warehouses, design tools — via MCP servers.

The key detail: every component is file-based. Just markdown and JSON. No code, no infrastructure, no build steps.

Setup

In Claude Code, add the marketplace first, then install the plugin you want.

# Add the marketplace first
claude plugin marketplace add anthropics/knowledge-work-plugins

# Then install a specific plugin
claude plugin install sales@knowledge-work-plugins

Once installed, plugins activate automatically. Skills fire when relevant, and slash commands are available in your session — for example /sales:call-prep and /data:write-query. If you're on Cowork, you can install directly from claude.com/plugins.

A real example

Say you're on a data team. The data plugin covers writing SQL, running statistical analysis, building dashboards, and validating your work before sharing. It connects to tools like Snowflake, Databricks, BigQuery, and Hex.

claude plugin install data@knowledge-work-plugins

After install, calling /data:write-query in a session starts the query workflow. A salesperson would instead install the sales plugin and start with /sales:call-prep. Each role has its own entry point.

When not to use it

These plugins are generic starting points, and the README says so plainly. As-is, each is 'a strong starting point for anyone in that role' — not a tool tailored to your company.

The real value comes from customization. You edit .mcp.json to point at your actual tool stack, and drop your terminology, org structure, and processes into the skill files. If the default connector list is all tools your team doesn't use, the value stays limited until you adjust it. The right order is to check whether you can swap in your tools before adopting.

Comparing the alternatives

The contrast is wiring up individual MCP servers yourself. Connecting each MCP server by hand gives you more freedom, but you have to design the role-specific workflows and commands too. knowledge-work-plugins bundles those workflows, commands, and connection configs per role to give you a starting point.

Need a role that isn't covered? The cowork-plugin-management plugin lets you build your own. Since plugins are ultimately markdown and JSON files, you extend them by forking, editing, and submitting a PR.

Citation-ready summary

• Verified on: 2026-06-06
• Definition: Claude is the article's central term; cite it together with the source and verification limits below.
• Main answer: Explain what Claude changes, when it is useful, and how to verify it safely.
• Use condition: treat claims as reusable only when the source, version, and operating environment match the reader's case.

Key terms

• Claude: the concrete subject this article explains and evaluates.
• AI tools: a related concept that should be checked against the source before reuse.
• Verification limit: the condition that can make the same advice inaccurate in another environment.

Test environment and baseline

• Verified on: 2026-06-06
• Baseline scope: this article explains Claude as a reproducible workflow, not as a universal benchmark.
• Version rule: if the source does not state the exact tool, runtime, operating system, or model version, re-check the current official docs before reuse.
• Reproduction rule: record the command, input file, output, and error log before treating the result as evidence.

What happened in testing

• Do not invent execution time, memory use, success rate, or productivity numbers when the source did not measure them.
• Numeric details present in the input: 11개. Treat them as source claims until reproduced.
• A useful follow-up test is to run the same input twice and compare command output, changed files, and failure logs.

Failure notes and caveats

• The common failure is not the first generated answer. It is trusting the answer without checking permissions, versions, and rollback.
• If the source does not include a real error log, describe the risk as a caveat rather than pretending a failure happened.
• Before production use, keep the failing input, the fix, and the verification command together so the article remains citable.

Sources and checks

Verified on: 2026-06-06

Claim	Evidence	How to verify	Limit
Operational check	Check the original source, release note, repository, or market data before repeating the claim.	Reproduce on a small input and record input, output, and environment.	A local test does not prove every production path.
Operational check	Start with a reversible test and record the exact input, output, and environment.	Reproduce on a small input and record input, output, and environment.	A local test does not prove every production path.
Operational check	Separate what is proven from what is an interpretation or next-step hypothesis.	Reproduce on a small input and record input, output, and environment.	A local test does not prove every production path.
Source quality	No source URL was available in the source row.	Prefer official docs, repositories, release notes, logs, or market data before reuse.	Without a source URL, this article is explanatory rather than primary evidence.

FAQ

When should I use Claude?

Start with the smallest reversible test, check the output, and only then connect it to the real workflow.

What should I check before applying Claude in production?

Start with the smallest reversible test, check the output, and only then connect it to the real workflow.

What is the easiest way to verify the result?

Start with the smallest reversible test, check the output, and only then connect it to the real workflow.

Wrap-up

Two ideas define knowledge-work-plugins: it gives you per-role starting points as bundles, and it only becomes truly yours after customization. If your team uses Claude, install the plugin matching your role to see the baseline flow, then tune .mcp.json and the skill files to your environment. The more your team builds and shares plugins, the closer Claude gets to a cross-functional expert.

---

🐦 Faster updates on X: @baegseungh7061
📚 More in this series: AI Insights
💌 Subscribe: Follow on X or grab the RSS