| AI Assistant in Onam Security | Available | Ask questions about your posture in plain language; specialists query your findings and cite them. Read-only. |
| Workspace | Early access | Ask Onam AI in plain language, each reply attributed to the agent that produced it, with a context panel for evidence, tasks and suggested fixes. |
| Orchestrator and visible plans | Early access | A multi-step question becomes a plan you can see — which agent, which skill, in which order — before and while it runs. |
| Asset Agent and Security Agent | Early access | Assets, relationships, crown jewels, findings, exposure and attack paths over the Onam Security inventory and findings, on AWS first. |
| Sensitive values masked | Early access | Answers name the resource and the finding; secrets, keys and sensitive data values are masked. |
| Evidence on every claim | Early access | Each number cites the skill, tool, query, rows and scan time that produced it. |
| Approval centre | Early access | One queue for every suggested change, diff first, with rollback and blast radius shown before the decision. |
| Activity and audit trail | Early access | Every turn, tool call, decision and approval recorded append-only, with a hash-chain check that nothing was edited. |
| Reports | Early access | A conversation or task exported with its evidence preserved. |
| Kill switch and autonomy ceiling | Early access | Stop all agent activity for your organisation; cap how far agents may go — read, analyse, recommend, simulate, propose. |
| Compliance Agent and Data Agent | In development | Built; waiting on the compliance-results, exposure and relationship feeds. |
| Change proposals with rollback | In development | Automation Planner proposals with exact change, rollback and dry run; blast-radius inputs are being connected. |
| Pre-built workflows | In development | Multi-step procedures such as audit preparation and exposure remediation, with approval steps built in. |
| Executing approved changes | On the roadmap | The actor applies an approved change in an isolated sandbox, validates it and rolls back on failure. Built and tested; enabled for no customer. |
| Azure, Google Cloud and Kubernetes actions | On the roadmap | AWS first; the design keeps provider specifics inside tools so agents do not change. |
| Event triggers and a workflow builder | On the roadmap | Start investigations from events; compose workflows that cannot publish an execute step without an approval step. |
| Customer-authored agents | On the roadmap | Needs a sandbox and certification model that does not exist yet. |