Skip to main content
Glama

skylight_upload_slots

Read-only

Generate signed upload URLs for up to 10 photos or videos so hosted deployments can upload files the server cannot read locally, then pass each returned blob key to photo tools.

Instructions

Hosted deployments (the claude.ai connector): get signed upload URLs so skylight_upload_photo / skylight_import_events_from_photo can take a photo or video the server cannot read from your disk. For each file returns {blobKey, uploadUrl, curl, expiresAt}: PUT the file to uploadUrl with exactly the Content-Type shown (the curl does it), then call the photo tool with image_blob_key set to blobKey. Types: jpg, jpeg, png, heic, gif, webp, mp4, mov (max 200 MiB). Writes nothing to Skylight; each URL is a credential to write that one upload and expires in 30 minutes. Refused when this server has no blob store (a local install: pass image_path instead). Up to 10 files per call.

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": {
      "files": "<files>"
    },
    "name": "skylight_upload_slots"
  }
}

Input Schema

NameRequiredDescriptionDefault
filesYesOne entry per file to upload: {filename, mimeType?}. The filename (a bare name, no path) is what Skylight will see; its extension sets the type when mimeType is omitted.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Addedv1.6.0

TDQS

A4.6/5.0
Behavior4/5

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

Annotations only say readOnlyHint and openWorldHint, but the description adds substantial context: nothing is written to Skylight, each URL is a single-use credential that expires in 30 minutes, the PUT must send exactly the returned Content-Type, and the call is refused without a blob store. The 'writes nothing to Skylight' claim is consistent with readOnlyHint, though the description could say more about failure/partial-batch behavior when one of up to 10 files is invalid.

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-loaded with the deployment context and the outcome, then the return shape, then the workflow, then constraints. Dense but nearly every clause carries operational weight; the parenthetical asides ('the curl does it', 'a local install: pass image_path instead') slightly interrupt the flow but are useful rather than filler.

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?

For a one-parameter tool with no output schema, the description covers the return fields ({blobKey, uploadUrl, curl, expiresAt}), the required follow-up call, the credential's expiry, size/type limits, and the refusal case. An agent has everything needed to execute the two-step upload correctly.

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 100%, so the baseline is 3, but the description adds meaning not in the schema: accepted types (jpg, jpeg, png, heic, gif, webp, mp4, mov) and a 200 MiB per-file cap. It restates the 10-file limit and the Content-Type requirement, but the size ceiling in particular is genuinely new information an agent needs before calling.

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 ('get signed upload URLs') plus the exact downstream tools it enables (skylight_upload_photo / skylight_import_events_from_photo) and the reason it exists (the server cannot read your disk). An agent can immediately tell this is the hosted-deployment file-ingest helper, distinct from the photo tool itself.

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

Usage Guidelines5/5

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

Gives explicit when-to-use (hosted/claude.ai connector), an explicit when-not and alternative ('Refused when this server has no blob store (a local install: pass image_path instead)'), and the exact follow-up step: PUT to uploadUrl, then call the photo tool with image_blob_key = blobKey. Routing between this tool and the direct-upload path is fully resolved.

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

Deploy Server

Other Tools