Filelayer
User files for a SaaS, under one person's authority: list, share with an expiry, revoke.
The current version, v0.33.0, was published to the official MCP registry on 2026-10-16. It is distributed as @filelayer/core on npm, filed under Other by this index, and declares 14 environment variables. Removals and unreachable sources are measured across every server this index tracks, contract drift across the servers that answer in consecutive snapshots; the index's current counts put this record in context.
Using Filelayer in Claude, Cursor, Gemini CLI, Cline, or Zed?
MCP tool contracts can change remotely with no version bump. The mcpindex gate pins each contract and HOLDs the call when it drifts-before your agent acts. Zero credentials. This is not the package install for this server itself (use Install this server for that).
Rewrites your MCP host config so each server launches behind the gate. Inspect first: curl -fsSL https://mcpindex.ai/install.sh | less
uv tool install mcpindex-gate && mcpindex-config-wireSemantic screen found no manipulation pattern in the description. Conformance probe not yet run.
mcpindex.integrity.descriptionpassINFOevidence“Normal file management capability”via static_description
- - Semantic screen only - the deterministic conformance probe has not run on this server
- - Confidence is reported but not yet calibrated (v1)
- - Screen reads the tool description, not the live behavior
- - advisory
- - registry description only no input schema
- - screen model 8b
- - The registry listing (version, packages, links) changed after this screen ran - the description we assessed did not
Semantic screen: an LLM judge reads the tool description for hidden instructions (status PARTIAL). A pass means the description is not lying, not that the tool is safe: a high-capability tool with an honest description still warrants caution. The deterministic conformance probe has not been run on this server yet, so the screen here is semantic-only. Posture: advisory. Confidences are reported but not yet calibrated (calibrated=false at v1). Full verdict history is not shown on this page.
Own this server? Test a description on the public screen → It is a separate check from the verdict above, so the two can differ.
That verdict was true at screening time (snapshot 2026-10-20).
Contracts can change after screening, with no version bump. The gate pins Filelayer’s tool contracts on first sight and holds any silent change before your agent acts - the check that keeps being true on Tuesday.
See your first HOLD in 2 minutes →
Related: how to trust an MCP server · screen before install · silent contract drift
People vet a server before wiring it in. One line on your project site or docs answers that with an independent record: screening verdict, registry provenance, contract drift history. It stays current as new screens and drift events land.
[Independent trust record for Filelayer](https://mcpindex.ai/server/dev-filelayer-core) - screening verdict, registry provenance, and contract drift monitoring.<a href="https://mcpindex.ai/server/dev-filelayer-core">Independent trust record for Filelayer</a> - screening verdict, registry provenance, and contract drift monitoring.A live verdict badge for your README or listing. It reflects the current screen, links back here, and updates when the verdict does.
[](https://mcpindex.ai/server/dev-filelayer-core)<a href="https://mcpindex.ai/server/dev-filelayer-core"><img src="https://mcpindex.ai/api/v1/badge/dev-filelayer-core" alt="mcpindex verdict" height="20" /></a>DATABASE_URLYour own PostgreSQL, where this library keeps file ownership, grants and the audit chain. Apply node_modules/@filelayer/core/schema.sql once before starting.
FILELAYER_ASThe ONE subject this server speaks for, in your own id space. Fixed for the life of the process: no tool reads it as an argument and no tool can change it, which is the only reason exposing these tools to an agent is safe. Write it yourself or have your provisioning write it, never derive it from a conversation. Serving several people means starting several servers.
FILELAYER_ORGThe organisation this server is scoped to, in your own id space.
FILELAYER_DATA_DIRA local directory for the bytes. Needs no bucket and no IAM user. Set this OR the S3_* block: with neither, the server refuses to start rather than falling back to memory, which would work in a demo and lose the first real file.
S3_ENDPOINTAn S3-compatible endpoint, for example https://<account>.r2.cloudflarestorage.com. Setting it makes S3_BUCKET, S3_ACCESS_KEY_ID and S3_SECRET_ACCESS_KEY required.
S3_BUCKETThe bucket. It must be private: this library decides who may be GIVEN a URL and has no say in who the object store will answer. `npx -p @filelayer/core filelayer doctor` reports whether yours serves unauthenticated reads.
S3_REGIONRegion, or `auto` for Cloudflare R2.
S3_ACCESS_KEY_IDAccess key id for the bucket.
S3_SECRET_ACCESS_KEYSecret access key for the bucket.
FILELAYER_MCP_AGENT_LABELHow this agent names itself in the audit trail, recorded on every event these tools produce. It is what lets the chain separate "the partner opened this document" from "the partner's assistant opened this document".
FILELAYER_MCP_ALLOW_DESTRUCTIVESet to exactly `true` to register delete_file. Off by default because deleting is not recoverable and an agent that misreads a sentence deletes the wrong thing silently.
FILELAYER_MCP_EXPOSE_AUDITSet to exactly `true` to register file_audit. Off by default because the trail is an organisation's whole history and a tool that returns it copies every file name, actor and denial into the model's context. Reading it is itself audited since 0.27.0, which was the previous reason; the default did not move.
FILELAYER_MCP_RETURN_BYTESSet to exactly `true` to register read_file. Off by default because a document returned by a tool is a document copied into a model's context; file_info plus a share link is usually what was actually needed.
FILELAYER_BASE_URLThe base of the share URLs this instance mints. Without it, share links come back without a URL.
Test ideas, messages and ads with Human Digital Twins built from real people, in your AI assistant.
Bet slip grader: grade a parlay A+ to F- before you bet, track slips, history and Tail. 21+, US.
Build on the Flarelink stack — Cloudflare auth, D1, and R2 — from your AI coding tools.