Security

Reporting a vulnerability

If you find a security problem in Sutra, tell us. This page sets out where to send it, what is in scope, and the dates we commit to.

How to report

Write to contact@optsutra.com with Security: at the start of the subject. A person reads every report. Please do not publish a proof of concept before we have agreed a date.

We publish no encryption key today. If you want an encrypted channel, send one line asking for it, with nothing sensitive in it, and we will agree one first.

What you tested

The build and the exact route, file or screen. Every page of this site carries its build in a sutra:version meta tag. For a deployed instance, ask the utility's administrator.

How it works

Steps we can follow, the smallest proof that the issue is real, and what an attacker gains. The last part sets severity.

Any data you reached

If you reached data that is not yours, say what and how much. The safe harbour below covers this. Tell us how you want to be credited, or that you prefer not to be.

Scope

In scope

The Sutra platform as deployed for a utility, only with that utility's written permission and inside the boundary they set. This website and anything served from it. Sutra source code, container images or builds shared with you under a pilot, audit or partner agreement.

Out of scope

Social engineering of our people, a customer or a customer's staff. Denial of service, load or stress testing, and anything that degrades a running utility. Physical access to a utility site, and attacks needing an already compromised device. Services we do not run: the code host, DNS and edge, mail, a model gateway. Scanner output with no demonstrated impact. Documented development defaults.

A deployed instance sits inside a utility's network, and only the utility can authorise testing it.

What happens after you report

Working days are Monday to Friday, India time.

3 working days

Acknowledgement

A reply from a person confirming we have your report.

10 working days

Verdict and severity

Whether we reproduced it, its severity, and the fix deadline. If we cannot decide, we say what is blocking us.

Every 14 days

Updates

A status update while the report is open.

Same day

Fix shipped

We tell you, and name the builds that carry the fix.

Each report ends in one of four states. Confirmed: reproduced, severity set, fix scheduled. Not reproduced: we say what we tried and what would settle it, and the report stays open for your reply. Out of scope: named against the list above. Accepted risk: real, not yet being fixed, with the reason written down. We never record "not reproduced" as "not a vulnerability".

Severity and fix targets

We set severity by impact. Fix clocks start on the day we confirm the report.

Critical, 7 days

Unauthenticated access to another tenant's data, remote code execution, authentication bypass, or a secret that unlocks any of those.

High, 30 days

Authenticated cross-tenant read or write, privilege escalation, query-language injection, or a control that shows as on while it is off.

Medium, 90 days

Exploitable only under unusual preconditions, or with limited reach.

Low, 180 days

Low impact, or hardening with a demonstrated path. Fixed within the window or in the next release, whichever is sooner.

If a fix will miss its target, we say so in writing with the reason and a new date, and offer a mitigation the utility can apply meanwhile.

Disclosure

We ask for 90 days from our acknowledgement before you publish. If the fix ships sooner, you may publish once affected deployments are updated. We extend only by written agreement, for a stated reason, usually a utility's maintenance window.

When a fix ships, we publish a short advisory: what was wrong, what an attacker could do, and which builds are affected. We credit you by name unless you ask us not to. We will never require a non-disclosure agreement as a condition of reporting.

Safe harbour

If you stay in scope, take only the data you need to show the issue, damage nothing, and give us the disclosure window, we treat your research as authorised. We will not bring or support legal action against you over it, and we will confirm this in writing to anyone who asks.

This covers systems we control. Only a utility can authorise testing of a Sutra instance inside its network. If you want to test a live deployment, tell us and we will ask the utility. Nothing here waives a third party's rights or overrides Indian law.

Deployed systems and CSIRT-Power

When a confirmed vulnerability affects a Sutra deployment, we tell that utility directly, with the severity, what an attacker can reach, and a mitigation to apply before the fix lands, whoever reported it.

Regulation 11(7) of the CEA (Cyber Security in Power Sector) Regulations, 2026, notified 31-Jul-2026 and in force from 1-Apr-2027, requires a vendor to run a structured process for reporting vulnerabilities and to pass them to CSIRT-Power. This page is that process, and we will report confirmed vulnerabilities in deployed systems to CSIRT-Power.

As of 6-Sep-2026 we hold no registration or direct channel with CSIRT-Power. Today the route is the affected utility's own reporting path plus CSIRT-Power's published contact, and opening a direct channel is an open item. This programme accepts reports from anyone, which is wider than the regulation requires. Incident reporting is a separate duty that belongs to the operating utility.

If in doubt, send it.

Sutra.
The Energy OS for India's power distribution, with every answer auditable.
livemodelledillustrative
Products
PulseRegulatory Benchmarkingruns todayAtlasroadmapProcurement & MarketsroadmapProoflisted utilities
Company
PlatformSolutionsIndia Energy StackReadiness mapWhy SutraCareersInvestors
Engage
How we engageBook a pilotFor DISCOMsFor regulatorsFor partnersFor private utilitiesContactSecurity