Feishu MCP Server
Infrastructure · tested 2026-08-22 · re-test due 2026-11-22 · by the Hlido desk, not the vendor
In short: An early, thin MCP server for automating Feishu (Lark) documents with AI — a real, buildable idea, but the public surface is a minimal no-code landing page with little to verify.
4 PASS · 1 FAIL of 5 public-surface claims
Quick answer
Feishu MCP Server scores 52/100 (FADING) on Hlido’s independent, hands-on test (reviewed 2026-08-22). FADING (52) is a low-confidence reading of a thin surface, not a verdict that the tool is broken. Pricing was not findable on the public surface when tested.
This project exposes an MCP server that lets an AI assistant operate on Feishu (Lark) documents — the user supplies a document link and an instruction, and the changes are made in the doc. The idea is legitimate and useful for the large Feishu/Lark user base, and the stated stack (Cloudflare Workers + Hono) is a sensible, low-cost way to ship an MCP endpoint. But the review has to be honest about what's actually on the public surface: a thin, Lovable-hosted landing page (fms-tapeless.lovable.app), mostly a single value-prop and a 'join our community' call, a use-cases page, and a 'view source' link — with no visible documentation depth, no pricing, no security/permission model for the Feishu access it would hold, and no evidence of adoption or maintenance. That's an early hobby-to-community-stage surface, not a productized offering, and the score reflects low confidence rather than a judgment that the underlying tool is bad. The concept could be genuinely handy; the public presentation gives an evaluator very little to stand on, and handing an MCP server write access to your Feishu documents warrants a permission model this surface doesn't describe.
Why FADING
FADING (52) is a low-confidence reading of a thin surface, not a verdict that the tool is broken. The concept — an MCP server for AI-driven Feishu document automation — is real and the stack is sensible, but the public presentation is a minimal Lovable landing page with no documentation depth, no security model, no pricing and no adoption signal, so almost nothing can be independently verified. Not lower because there is a genuine, working-sounding product idea and a source link; not higher because the evidence surface is too thin to support a confident STEADY.
Public-surface checklist
- PASS Homepage loads (required)
- PASS Primary value prop (required) — Feishu MCP Server — connect AI assistants to Feishu, no coding
- PASS Cta present — 现在体验 / 立即体验 (Try now)
- FAIL Pricing or access — No pricing or access terms on surface
- PASS Evidence or demo — Use-cases page; 2 screenshots captured
What we saw
2 screenshots captured by the Hlido engine during the reviewed run (run-5a93d917118a27f6-fms-tapeless-lovable-app). Our own captures — not vendor marketing material.
What it does well
- Solves a real need: AI-driven automation of Feishu (Lark) documents via MCP
- Sensible, low-cost architecture stated (Cloudflare Workers + Hono)
- No-code framing lowers the barrier for non-technical Feishu users
- Source link offered, allowing some inspection of the actual implementation
What it fails at
- Public surface is a thin Lovable-hosted landing page — no real documentation depth
- No security/permission model described for the Feishu document access it would require
- No pricing, no adoption signal, and no evidence of active maintenance on the surface
- Product maturity reads as early/community-stage rather than a dependable offering
Red flags
- An MCP server that edits your Feishu documents needs a clear permission and data-handling model; none is described on the public surface. Inspect the source and scope access carefully before use.
- Hosted on a generic Lovable subdomain with a 'join our community' framing and no maintenance/adoption signal — treat as early-stage until proven otherwise.
Best for
- Feishu/Lark users experimenting with AI-driven document automation
- Developers comfortable reading the source and self-hosting an early MCP server
- Tinkerers who want a low-stakes way to try MCP against Feishu docs
Not recommended for
- Teams needing a documented, supported, security-reviewed integration
- Anyone who requires a clear permission/data-handling model before granting document access
- Production use where reliability, maintenance and support guarantees matter
Pricing & access
- Pricing findable on the public surfaceFAIL No pricing or access terms on surface (tested 2026-08-22)
Derived from Hlido-held evidence only (engine checklist + editorial text); quotes are verbatim from the scorecard; not vendor-supplied; re-derived daily. Verify current prices on the vendor's pricing page. Last verified 2026-08-22.
Related agents
Agent relevance
MCP
Agentic-Commerce Readiness 41/100 · SURFACE-ONLY
Independent readiness for agent delegation & transaction. How it’s scored · check live
The product is itself an MCP server, so an AI assistant can in principle connect to it and act on Feishu documents. In practice the public surface documents neither the endpoint, the tool schema, nor the auth model, so integration would require reading the source; it is not verifiable from the marketing surface.
Agent-friendly score: 4/10
Score over time
The longitudinal record — every point is the score as published on that date. Raw series.
Evidence
scorecard.json · transparency passport · registry · methodology
Embed this trust badge
Markdown
[](https://hlido.eu/check/?agent=ztxtxwd-open-feishu-mcp-server)HTML
<a href="https://hlido.eu/check/?agent=ztxtxwd-open-feishu-mcp-server"><img src="https://hlido.eu/badge/ztxtxwd-open-feishu-mcp-server.svg" alt="Hlido trust score"></a>