Agents

Who it is for

Automation engineers who need dry-runs and human gates before production execution.

What works today

• ·Automations console with dry-run API
• ·Robot backend proxy when configured
• ·Remediation receipts and approval linkage

Problem

Unsupervised automation in production creates blast-radius risk without dry-runs or human gates.

Solution

Agents extend Platform automations with playbook execution against robot backends, policy guardrails, and audit receipts.

How it works

1. Define playbooks in the Automations console.
2. Dry-run records results before live execution (paid plans).
3. Robot backend reached via server proxy when configured.
4. Remediation flows capture steps and blocked reasons.

Benefits

• ·Prototype execution path exists today in Platform
• ·Clear simulated vs connector modes
• ·Tied to incidents and approval notes

Roadmap

• →First-class agent registry and scheduling
• →Memory Pendant context injection
• →Multi-step agent graphs