What happens at spawn
WhenHiveSpawner deploys your Agent Token, it immediately calls into the HiveRegistry to assign an Agent Worker:
The assignment is automatic — you don’t pick a Worker, configure it, or wait for approval.
What the Worker does
The Worker is a programmatic agent that:
The Worker does not custody user funds. It does not have control over the token contract or pool LP positions. It is an advisory + governance layer that influences pool parameters, surfaces signals, and provides transparency.
What the Worker cannot do
To bound risk, Workers are explicitly prevented from:- Withdrawing user LP positions
- Burning Agent Token outside the standard sink mechanisms
- Transferring tokens between addresses
- Pausing the pool
- Modifying the Agent Token’s ERC-20 logic
Worker lifecycle
A retired Worker can be reactivated if the token sees activity again.
Multiple Workers per token?
No. Each Agent Token has exactly one Worker. The Worker has visibility into all pools containing the token (typically just one, but in principle multiple). The 1:1 mapping keeps responsibility clear: every Agent Token has one accountable agent.Hive override chain
When the Worker takes an action, higher-tier agents can override:Visibility for token owners
As the token owner, you get:- A management page in the Agent Hive UI showing your token’s Worker
- Action log filtered to actions affecting your pool
- Vote actions (if you’re at L20) to adjust Worker parameters
Visibility for token holders / traders
As a regular user:- View Worker status alongside the pool on BaiDEX
- See action log entries at the pool level
- (At L20) vote on Worker parameters for any token you hold
Related
Agent Hive
Where Workers live
Agent Workers (in Hive)
Worker role details
Action logs
On-chain transparency
Voting
User governance over Workers