Skip to main content
Glama

Save a Tunc Art file

art_write

Save a file for the user to keep, open or share: a report, export, image, spreadsheet or document, any type, up to 250 MiB. Use it when the user asks you to produce or store a file; the result gives you its private link to hand back. The link is private to the user; they can share the file with anyone by setting a password on its page. Set mime_type to the file's real media type so its link previews (images, video, audio, PDF, HTML, text, Markdown, JSON, CSV and code); other types download. Left out, the type is worked out from the file's bytes and name. Send small files inline (content_base64 or text, up to 8 MiB); for larger files use mode="begin" to get an upload URL, PUT the bytes there, then mode="complete" with the upload_id. The result's artifact carries the file's id (its artifact_id) and viewer_url, its private link. Files are kept for the account's retention period (each file shows its expires_at; longer with paid storage). When storage is full, the oldest files are removed first and the account lists what was removed. Keep the artifact_id to open the file later with art_read or remove it with art_delete.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "operation": "<operation>"
    },
    "name": "art_write"
  }
}

Input Schema

NameRequiredDescriptionDefault
modeNoinline (default): send the file in this call. begin: get an upload URL for a larger file (name, byte_count; optional mime_type, sha256). complete: finish an upload after the PUT (upload_id, byte_count; optional sha256).
nameNoFile name shown to the user, such as q4-report.pdf (1-240 characters). Required for inline and begin.
textNoinline only: the file as UTF-8 text (at most 8 MiB). Exactly one of content_base64 or text.
sha256NoOptional: the file's SHA-256 (64 hex characters), checked when the upload completes.
mime_typeNoMedia type, such as application/pdf or image/png. Set the real type so the file previews. Left out, it is worked out from the file's bytes and name (pictures, PDF, Office files, audio, video, archives and text are recognized); a file nothing identifies is application/octet-stream, which downloads.
operationYescreate: save a new file.
upload_idNocomplete only: the upload_id that mode=begin returned.
byte_countNobegin and complete: the file's exact size in bytes.
content_base64Noinline only: the file's bytes, base64 (at most 8 MiB decoded). Exactly one of content_base64 or text.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • changedInput schema / properties / mime_type / description
      Previous value: -"Media type, such as application/pdf or image/png. Set the real type so the file previews; the default (text/plain for text, application/octet-stream otherwise) previews only text, and application/octet-stream downloads."New value: +"Media type, such as application/pdf or image/png. Set the real type so the file previews. Left out, it is worked out from the file's bytes and name (pictures, PDF, Office files, audio, video, archives and text are recognized); a file nothing identifies is application/octet-stream, which downloads."
  2. First observed

TDQS

A4.6/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Goes well beyond the annotations: size caps (250 MiB total, 8 MiB inline), the private-link ownership model, password-based sharing, preview-vs-download behavior by media type, the three-step begin/PUT/complete flow, retention and expires_at semantics, and the eviction policy when storage is full. These are exactly the non-obvious traits an agent needs before invoking a write tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loads purpose and trigger in the first sentence, then moves through media type, upload modes, result shape, and retention in a logical order. It is dense — a single long paragraph — but nearly every clause carries actionable information; only the sharing/password note is arguably peripheral.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

There is no output schema, yet the description explains the return payload (artifact_id, viewer_url, expires_at) and how to use it downstream. Combined with upload mode mechanics and limits, an agent has everything needed to call this correctly for both small and large files.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is already 100%, so the baseline is 3, but the description adds real meaning: the mode lifecycle (begin returns an upload URL, complete needs upload_id), why mime_type should reflect the real type, and how type inference falls back when it is omitted. It stops short of describing every parameter (sha256, byte_count) in the prose, leaving those to the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Save a file for the user to keep, open or share') and enumerates the concrete artifact kinds (report, export, image, spreadsheet, document). It also names the sibling tools art_read and art_delete, so an agent can distinguish save/create from open/delete without consulting schemas.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives an explicit trigger ('Use it when the user asks you to produce or store a file') and routes later lifecycle steps to art_read and art_delete. It does not state when NOT to use it (e.g., when the user just wants text in chat rather than a stored file), so it falls short of full when/when-not coverage.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources