What is sent
ArcRoom sends metadata about a session: what was asked, which tools ran on which paths, and how it ended. It never uploads your files or edit diffs on its own; of tool output, only a failed command's first 500 characters of error text go, so the next agent is warned. A file's text leaves only if you paste it into a prompt or your agent saves it to team memory. Every field below is redacted for secrets on your laptop before it is queued, and again before it is uploaded.
What leaves your laptop
| Prompt | What you typed, with the agent’s system reminders removed and secrets redacted. The first 1,000 characters are stored; the first 300 are sent live for recall. |
| Tool calls | The tool’s name and its file paths, search patterns, shell commands (first 1,000 characters, redacted), URLs and web queries, and the tool’s short description. |
| Where it ran | Machine name and folder path; at session start the project name, git branch, model name and the path of the agent’s transcript file (not its contents). |
| Done-checks | The check commands the agent ran, whether they passed, and how the turn ended. |
| Files touched | Git hashes of the files a turn read or wrote, so stale steps can be spotted. |
| Token counts | Tokens used per turn, read from the agent’s transcript on your laptop. |
| What ArcRoom handed over | The size of each piece of context ArcRoom added, and the text of hand-overs (up to 2,000 characters). |
| What agents save on purpose | Whatever an agent sends to ArcRoom’s own tools (remember, complete_task and the others): their input and output. |
| Failed commands | When a command fails: the command and the first 500 characters of its error (redacted), and what worked instead, so the next agent is warned before it repeats it. |
| Code map | Your repo’s folder and file paths with their git hashes, and which files your agent is editing right now (with your name and branch). Shared with your workspace. On by default. arcroom code off stops it for a repo. |
What never leaves on its own
- File contents, whether read or written
- Tool output, except the error text of a failed command
- Edit diffs
- Prompts given to sub-agents
- Your AI provider account and its keys
Two exceptions: text your agent saves to team memory with ArcRoom's tools (remember, complete_task) is sent, even if it quotes a file, and so is anything you paste into a prompt. Every saved memory is listed on the Memory page.
A rules file you imported is the other exception: when it changes on your main branch, its new lines are sent as rule cards to accept on the Rules page. Stop it by deleting the card.
That includes the contents of .env files, keys and credentials: the plugin sends no file's contents on its own. A path to such a file, or a command that names one, is sent only as [secret file].
See it yourself
Run arcroom sent to see what this laptop sent today: how many prompts and tool calls, how many values were redacted, and the latest events. arcroom sent 2026-10-30 shows one day, and adding --all prints every event exactly as it was sent. The last 7 days are kept in ~/.arcroom/sent/. What your agent saves to team memory is listed on the Memory page.
Events wait in ~/.arcroom/spool.jsonl for a few seconds before they are sent, so that file is usually empty.
What redaction removes
Before an event is queued, again before it is uploaded, and in every request the plugin sends:
| This machine’s secrets | The exact values of your secret-named environment variables (…_KEY, …_TOKEN, …_SECRET, passwords) and the secret-looking values in the project’s .env files, wherever they appear and whatever their shape. |
| Known key formats | Anthropic, OpenAI, GitHub, GitLab, AWS, Google, Stripe, Slack, npm, Hugging Face, Shopify, Twilio, SendGrid, Mailgun, DigitalOcean, Linear, PyPI, Doppler, Postman and Databricks keys, JWTs and private key blocks. |
| Secrets by name | name=value and "name": "value" pairs whose name says password, secret, token or key, and any field under such a name. |
| In URLs and headers | Passwords inside URLs (https://user:password@host), Bearer and Basic auth headers, cookies, signed-URL signatures. |
| On commands | Secret flags such as --password, --token, --api-key, mysql -p and curl -u user:password. |
It is best effort. A secret in a shape none of these match can still get through, so please don't paste secrets into prompts.
Send less
arcroom off sends nothing at all for a repo, on your laptop. See Isolation and controls. The full data flow, retention and infrastructure are on the security page.