TRUSCOR
Had an AI incident? Get rated
Solutions · Teams scaling agents

Approve write access with evidence.

For companies of 200+ staff with agents in CI/CD, payments and data. Know what the agent can reach before it gets write access, and keep a record of what it does.

01Reach

Map everything the agent can touch.

Tools, MCP servers, plugins and data flows, mapped from the inside, with identity, permissions and tool-use boundaries tested.

02Blast radius

See how far one failure travels.

We attack the system on purpose, trace every path a failure could take and estimate what it could cost. Some boundaries hold. The report names the ones that don’t.

03Record

Then record what it actually does.

SOVA Black Box runs in your cloud, VPC or on-premises, with your keys. Every action is signed and checked against what took effect. If something fails, Incident forensics reconstructs what happened and what it cost.

Year-one package

Production Assurance

Complete rating on two orchestrations, with SOVA Black Box and Continuous Rating on both for 10 months.

$15,080 ₹5,77,000 · Year one
Prices exclude GST · Price depends on scope, never on the score
Or start with · Complete Rating, then SOVA Black Box
Complete Rating$2,800₹95,000Everything in Standard, plus design review, certification status, Evidence Pack and one re-rating
SOVA Black Box$300/mo₹12,000/moPer orchestration · $600 / ₹25,000 one-time setup
Forensics retainer$300/mo₹10,000/moOn call with a 4-hour response time
The job you’re hiring us for

“Before our agent gets write access, I want to know what it can reach and have a record of what it does, so I can approve it and prove it later.”

For agents in production

Measure the controls, record the actions.

TRUSCOR measures controls and never operates them. It tests whether a kill switch works but never supplies or pulls one.

The SOVA stackHighlighted: the parts behind this page
Effect-verified recording SOVA Black Box · O.W.L

Read-only O.W.L sensors check each recorded action against the real state of files, processes, network, browser, screen and databases, and mark it verified, failed or unknown.

Available
Oversight drills O.W.L · SOVA Engine · SOVA Black Box

We pull the kill switch inside an O.W.L test environment and measure time to halt, whether sub-agents and in-flight tool calls stop, and what state is left behind. We seed known errors into human review queues and measure how often reviewers catch them.

Available
Continuous Rating TAFAAR · SOVA Black Box

A monthly rating from the record, flagged “retest due” when the system’s behavioural fingerprint changes. ₹8,000/month.

Available
Self-expiring ratings TAFAAR · SOVA Black Box

Every grade is bound to a behavioural fingerprint of the whole system: model, prompt, retrieval and tools. When the fingerprint changes significantly, the grade flips to “stale, retest due”, like a credit watch.

Available
Discovery-and-cure ledger SOVA Black Box · O.W.L replay

Timestamps when our testing found a problem, and records an independent retest showing the fix held. We sell the test and the retest, never the fix.

Available

Free continuous regression testing of your own agents on every model update or deploy, and a safe rehearsal of an agent’s real task in a cloned environment.

Available

Decoy credentials, documents, tools and planted memories; any touch is recorded as attributed evidence. Hash-chained memory diffs trace an incident to the memory write that caused it. Detect and record only, never block.

Available
Evidence freezer SOVA Black Box · .sova · O.W.L

One action captures the full incident state (prompts, retrieved context, model fingerprint, memory, tool calls, credentials, approvals and downstream actions), sealed in .sova under TRUSCOR’s countersignature and replayable in O.W.L.

Available
Being built = the TRUSCOR corpus and its data products, which grow with every consented rating.

Get rated.

Five fields and a short call. We reply within one working day.