File storage for AI agents
Workspace files your agent can list, read and write. Chat uploads, skill outputs and files pushed over the API all land in the same place.
Copy the setup prompt.
Your agent does the rest.
Fetch and execute the appropriate instructions to set me up for superagnt_ from https://mcp.superagnt.com/agent-setup/prompt.md» On by default. The agnt_files_* tools are part of the base tool surface, so a fresh MCP connection already has them.
Everything your agent can do with it
agnt_files_list pages newest first and filters by source, mime prefix or name prefix.
agnt_files_get returns name, mime type, size, kind, source and status, so an agent can decide whether a file is worth reading.
agnt_files_read returns utf-8 lines or a byte slice for text, and base64 for images, PDFs and anything else.
A 256 KB cap per call with offset and limit, so a large file is read in pieces instead of flooding a context window.
agnt_files_upload publishes a report, an export CSV or a generated PDF, visible to you, to other agents, and in a later session.
Chat uploads, skill outputs and API uploads are the same files with a source tag, so nothing is stranded in one channel.
tool names04as your agent sees them
agnt_files_listagnt_files_getagnt_files_readagnt_files_uploadThree jobs it can start on today
Drop a file in chat, act on it
You upload a CSV in a conversation. The agent lists workspace files, reads the newest chat upload, and loads it into the workspace database without you moving it anywhere.
Deliverables that outlive the chat
A run produces a report. The agent uploads it as a file, so it is downloadable later and readable by the next agent instead of buried in a transcript.
Attachment triage
The agent filters files by mime prefix, reads new PDFs as base64 for extraction, and writes the extracted fields to a table.
What comes back
Your agent reads this response directly, with no parsing layer of your own in between.
https://mcp.superagnt.com/mcptool callagnt_files_list{
"source": "chat-upload",
"mime_prefix": "text/",
"limit": 2
}{
"files": [
{
"id": "file_7c1e",
"file_name": "q3-target-accounts.csv",
"mime_type": "text/csv",
"byte_size": 184320,
"source": "chat-upload",
"created_at": "2026-08-28T11:04:19Z"
},
{ "...": "1 more file" }
],
"next_cursor": "eyJjIjoiMjAyNi0wOC0yN1QxNjo0MTowMloifQ"
}What it costs to run
Platform tools carry no per-tool price. What you pay is the AI and credit usage the work causes, on the same workspace balance every other tool draws from.
Each run is itemized, so you can see which agent spent what and on which call.
how credits work →Before you wire it up
01What can it read?
Text and CSV come back as utf-8 lines or a byte slice. Anything else, including images and PDFs, comes back base64 encoded.
02How large a file can it handle?
Reads are capped at 256 KB per call. Pass offset and limit to walk a larger file in pieces.
03Where do the files come from?
Chat uploads, skill outputs and files pushed through the API all land in the same workspace store, tagged with their source so a filter can narrow to one of them.
04Do I need to enable it?
No. File tools are in the default tool surface for a workspace MCP connection at https://mcp.superagnt.com/mcp.
Point your client at https://mcp.superagnt.com/mcp and your agent has this, plus everything else on one balance.