> ## Documentation Index > Fetch the complete documentation index at: https://docs.attensira.com/llms.txt > Use this file to discover all available pages before exploring further. # Approval modes > One workspace setting decides how much the agent does on its own before anything leaves the workspace — Ask, Draft or Act. ## One setting, every automation and chat Approval mode is a single workspace setting: **Ask**, **Draft**, or **Act**. It applies to every [automation](/agent/automations) run and every [chat](/agent/chats) session in the workspace — there is no per-automation override. ## Ask The agent stops and asks before anything that leaves the workspace — publishing a page, sending mail, spending with a third party. A card shows exactly what it wants to do. The run then waits for you, for up to **7 days**. If nothing is decided in that window, the run closes without doing the thing, and files a row in your [Inbox](/agent/inbox) saying so. See [what the agent shows you](/agent/what-the-agent-shows). ## Draft (default) The agent does everything inside the workspace on its own, and files everything else as work awaiting your approval. Nothing is published, sent, or written outside the workspace until you approve it. This is the mode every workspace starts in. ## Act The agent carries **safe changes** through on its own, without stopping. Safe means hygiene work: structured data, meta descriptions, alt text, internal links, heading structure. Anything touching a claim — a number, a price, a product or competitor name, a commitment — is filed for your approval **in every mode**, Act included. Code blocks and code samples are never edited, in any mode. ## What Attensira changes on its own, and what it always asks about Every proposed change is put into one of three classes before anything happens to it. The class is worked out from the change itself — the before-and-after of the file, and where it lives — never decided by the agent about its own work. JSON-LD and FAQ structured-data blocks; meta descriptions and `og:`/`twitter:` tags; image alt text; internal links to your own site; heading levels (not heading words); whitespace. A change is safe only if **every line it touches** is on this list. One line that is not makes the whole change a claim. Every structured-data addition carries the note: "hygiene, not a lever: we do not expect this alone to move an answer." Any added or removed line containing: a price or currency amount; a percentage; a number with a unit ("40 ms", "50,000 events per second"); a promise word (guarantee, guaranteed, always, never, uptime, SLA, refund, unlimited, free, fastest, best, leading, "#1", a "99." figure); a competitor's name; one of your product names; rewritten heading words; a rating, review count or price inside structured data; anything that is a sentence rather than markup. Wrapping does not help — a price inside a schema block is still a price. Approving a claim publishes it, whatever the mode. Code in fenced blocks, `
` or ``; source files (`.js`, `.ts`, `.tsx`, `.go`, `.py` and the rest); CI configuration, workflows, dot-directories; lockfiles and dependency manifests; dotfiles; anything outside the directories you have said hold your pages.

    A proposal that turns out to be one of these is refused, and a blocked row in the [Inbox](/agent/inbox) names the rule that stopped it.
  


**Editable directories** default to `/docs/**` and `/content/**`, plus any content roots you named when connecting the repository, plus `llms.txt`, `robots.txt` and sitemaps at the root. Changeable in project settings. Every change is recorded with who and when.

Every decision carries the rule that decided it, in plain words — the same rule you see above, not a paraphrase of it. The rules are scored against a set of 42 adversarial changes on every change to the classifier; the current miss rate is 0%.

## What no mode changes

Three things hold regardless of the mode you pick:

* We never merge a [pull request](/agent/github). A pull request is a proposal; closing it undoes it.
* We never delete anything irreversibly without an explicit yes.
* We never send mail from your domain on your behalf without one.

## Changing your mind is retroactive

Mode is checked again when the work actually runs, not only when it was proposed. If you tighten the setting between a chat suggesting something and an automation acting on it, the tighter rule applies at run time.

Your own approval of one specific change still goes ahead even if you have not changed the mode — approving an action and changing the workspace default are different decisions.

Every change to the setting is recorded: who changed it, and when.

## Why your workspace is in Draft

Workspaces created before September 2026 had a setting with this same name, but it was only a default applied to new automations — it did not govern outward writes. As of this change it governs real permissions, so every existing workspace was moved to **Draft** and its owner was emailed about the change.

If you want the agent to carry hygiene changes through on its own once the classifier ships, choose **Act** again — Draft does not revert on its own.