Frequently asked questions

A decision inbox for autonomous AI agents. Your agents do work unattended and queue the calls only a human should make; you clear them, and the decisions flow back to the agents, machine-readable.

In a live chat you're already there — you can see the image and answer inline, so piling would be friction. Pileless earns its keep when the human isn't in the loop: an agent runs for hours overnight and you wake to a sorted queue instead of a transcript you have to babysit.

No. Pileless is for productive agents — Claude Code, Codex, any model on your key driving it over MCP/API. It runs no AI of its own.

Add the Pileless MCP server to your agent's config. See the Claude Code and Codex guides — about two minutes each.

Any MCP-capable agent. We have step-by-step guides for Claude Code and Codex; the same server works with Cursor, Cline, and anything that speaks MCP. You can also call the HTTP API directly.

Create a pile of decisions, append items to an open pile, wait for your decisions (long-poll), check status, and attach files. See the tool reference on the Claude Code guide.

Open your inbox. Each tile is a decision — approve, reject, pick one, rank, edit, or answer, right in place. Ctrl+scroll to zoom from a dense list to a full reader; arrange by urgency, sender, or time, or drag your own order.

No. Pileless runs no language models on your content. Your own agent (your key) does the thinking. Your content is processed only to route, render, and validate — never used for training, mining, or sale.

With Bring-Your-Own-Cloud, your files live in your own Cloudflare R2 bucket — never on our servers. The decision text/metadata lives in our encrypted database, with strict per-workspace isolation. Full security brief →

We're built to prevent it: files stay in your cloud, your workspace is walled off from every other customer, access is least-privilege and redacted server-side, share links are scoped/expiring/revocable, and every access is logged in a tamper-evident trail. See Security for specifics.

Not yet — it's on the roadmap, and we say so plainly. We're happy to do a security walkthrough or sign an NDA/DPA with you directly in the meantime.

Yes. It's your data — you can delete your piles, and with BYOC a decision's files live in a bucket you own and control directly. Note: the append-only audit log keeps a tamper-evident record of that actions happened (not your file contents) per our retention policy.

For BYOC, yes — you create your own Cloudflare R2 bucket (free tier: 10GB, no egress fees) and give Pileless a scoped API token. Your files then live in your account. The R2 walkthrough covers every step. (Without BYOC, files sit on Pileless's encrypted storage.)

Your own. Your own Cloudflare account, your own bucket, your own API token. We store only the encrypted credentials and reach your bucket via short-lived links. You can revoke access anytime.