English | 中文

awesome-mcp-workflows

Find MCP workflow ideas and server patterns that hold up under product scrutiny, not just demo energy.

awesome-mcp-workflows is a bilingual, pattern-first repository for builders who need launchable ideas, credible workflow shapes, and honest editorial labels instead of fake leaderboards.

Quick Proof

• Three browsable catalogs: servers, workflows, and recipes.
• An explicit evaluation rubric, taxonomy, and submission template so the list stays consistent.
• Bilingual coverage from day one, so entries are useful in both English and Chinese-speaking workflows.
• Clear evidence labels like Pattern seed and Documented example to separate what is documented from what is inferred.

Start Here

• Want reusable MCP server shapes? Start with lists/servers.md.
• Want practical workflow ideas? Start with lists/workflows.md.
• Want combinations closer to a shippable product? Start with lists/recipes.md.
• Want to contribute without lowering the bar? Read docs/evaluation-rubric.md and docs/submission-template.md.

If you want a practical MCP idea catalog that filters for usefulness over hype, star the repo and use it as your working shortlist.

Editorial Position

This repo is intentionally honest about what it is:

• an editorial catalog, not a live leaderboard
• a place for reusable patterns, not hype-driven demos
• a launch-research aid for builders, not a source of fake market claims

Most entries start as Pattern seed content. When a contributor adds a specific tool, deployment, or public example, the entry should say what is documented and what is only inferred.

Quick Navigation

• Explore the catalog: lists/servers.md, lists/workflows.md, lists/recipes.md
• Understand the editorial rules: docs/evaluation-rubric.md, docs/taxonomy.md, docs/submission-template.md
• Contribute safely: CONTRIBUTING.md, CODE_OF_CONDUCT.md, SECURITY.md, docs/releases/v0.1.0.md

What Belongs Here

• concrete workflows with a clear user, trigger, and output
• server patterns with explicit trust boundaries
• combinations of servers and workflows that could plausibly ship as a product or internal tool
• submission guidance that helps contributors stay consistent

What Does Not

• vague "agent does everything" ideas
• unverifiable claims about adoption, maintenance, or performance
• copied vendor positioning with no workflow detail
• lists that optimize for novelty instead of usefulness

Repository Map

Catalog lists:

• lists/servers.md: reusable MCP server patterns
• lists/workflows.md: practical workflow patterns
• lists/recipes.md: stack combinations and launch angles

Editorial docs:

• docs/evaluation-rubric.md: scoring dimensions for entries
• docs/taxonomy.md: shared labels for maturity, trust, and domains
• docs/submission-template.md: copy-paste template for new entries

Collaboration and governance:

• CONTRIBUTING.md: contribution rules and PR checklist
• CODE_OF_CONDUCT.md: participation expectations and reporting path
• SECURITY.md: how to report security problems in repo content or tooling
• docs/releases/v0.1.0.md: draft notes for the initial public release

Entry Model

Good entries answer most of these questions in a compact, scannable way:

• What problem does this solve?
• Who is the primary user?
• What are the inputs and outputs?
• Which servers or capabilities are required?
• Where is the trust boundary?
• Why is this useful beyond a demo?
• What is known, and what is only a pattern-level suggestion?

Editorial Labels

• Pattern seed: a reusable shape or workflow idea without a specific public implementation attached
• Documented example: a pattern tied to a specific source or implementation note

Use the label that matches the evidence. Do not upgrade an entry because it "sounds production-ready."

Evaluation Lens

Entries are judged on a small set of practical questions:

• Is the job to be done concrete?
• Is the user pull believable?
• Are the trust and safety boundaries obvious?
• Is the operating model realistic?
• Is the pattern composable with other servers?
• Can someone understand or demo it quickly?

The full scoring guide lives in docs/evaluation-rubric.md.

Start Here

• Looking for server ideas: start with lists/servers.md
• Looking for agent workflows: start with lists/workflows.md
• Looking for launchable bundles: start with lists/recipes.md
• Want to contribute: read CONTRIBUTING.md and use docs/submission-template.md

Language Policy

• English is the primary structure for list content.
• Chinese readers should be able to understand the repo from README.zh-CN.md.
• Supporting docs may stay English-first if they are short and mechanical.

Contributing

Contributions are welcome, but the quality bar is higher than "interesting link":

• prefer workflow clarity over breadth
• state the evidence level honestly
• explain why a builder should care
• make trust boundaries explicit

See CONTRIBUTING.md.