Trust, safety & autonomy

HiveBase does real work for you — drafting, triaging, following up, acting. The most important thing to understand is how it stays safe. This page is the honest answer to the question every founder asks: "What will it do without me, and can I undo it?"

The one principle

Reads are free and automatic. Anything that changes the world outside HiveBase asks first — unless you've explicitly told it not to. Every gate below is a consequence of that single rule.

1. 01Agent wants to actAny proposed step
2. 02Outside world?No → often automatic · receipted
3. 03If outboundSend · post · write · push
4. 04Reversible?Still hold for approval by default
5. 05IrreversibleAlways confirm · show exact impact

1. 01

   Agent wants to act

   Any proposed step

2. 02

   Outside world?

   No → often automatic · receipted

3. 03

   If outbound

   Send · post · write · push

4. 04

   Reversible?

   Still hold for approval by default

5. 05

   Irreversible

   Always confirm · show exact impact

What's automatic vs. what asks first

HiveBase classifies every action by how reversible and how consequential it is.

Action type	Examples	Default behavior
Read	Search the Brain, summarize a thread, prep a brief	Automatic
Internal write	Record a decision, create a task, draft content	Often automatic · receipted
External / outbound	Send an email, post to Slack, update a CRM, push code	Asks first
Irreversible / high-impact	Bulk send, delete, anything you can't take back	Always confirms

See the plan before it runs

For multi-step work, HiveBase shows you the plan first — the list of steps it intends to take — so you can review, edit, or reorder before anything happens.

1. 01Proposed planSteps visible first
2. 02You review / editReorder · rewrite · drop
3. 03ApproveOr send back
4. 04Run stepsInternal continues
5. 05Outbound still holdsApprove each external act

1. 01

   Proposed plan

   Steps visible first

2. 02

   You review / edit

   Reorder · rewrite · drop

3. 03

   Approve

   Or send back

4. 04

   Run steps

   Internal continues

5. 05

   Outbound still holds

   Approve each external act

Stop it any time

• Every running task has a visible Stop control — one click halts it immediately.
• In-flight outbound actions hold at the approval gate; they never fire while you're deciding.
• Pausing never loses work — the task keeps its context and you can resume or redirect.

Everything is logged and reversible

• Internal changes (decisions, tasks, context) keep full history — nothing is overwritten.
• Every applied agent change leaves a receipt: what changed, when, and which sources supported it.
• Outbound actions you approved are logged with the exact content that was sent.
• When two sources conflict, HiveBase holds the contradiction for review instead of guessing.

What HiveBase will never do without you

• Send an email, message, or anything to a customer without your approval.
• Make an irreversible or destructive change silently.
• Act outside the integrations and scopes you've connected.
• Train external models on your private company data.

Permissions & data access

HiveBase only sees what you connect, and each connection starts read-only. Write-back to a tool (e.g. updating a CRM record) is a separate, explicit opt-in. Access is scoped to the integrations you enable — see Integrations and, for programmatic access, the developer scopes.

Layer	Control
What HiveBase sees	Only the tools you connect, read-only by default
What it can change	Internal by default; tool write-back is opt-in
Who sees what	Your team permissions + per-integration scope
Programmatic access	Per-key scopes (hb_sk_) — least privilege

Data retention & deletion

• Disconnect any integration at any time; HiveBase stops reading it immediately.
• Delete captured data; deletions are honored and propagate.
• Connected-tool tokens are revoked on disconnect.
• Your private data is never used to train external models.

Frequently asked questions

Related