Skip to content

For teams putting AI agents to work

Trace Every AI Action Make Compliance Verifiable

Every agent action is checked against your rules before it runs. Risky ones wait for your approval. Your auditors get proof, and your data never leaves your company.

  • Runs in your environment
  • Works with your agent framework
  • Built at NTU Singapore
Example: an AI agent's console. TraceStone checks each action against your rules before it runs. READ revenue is allowed. DELETE customer_records is blocked by the rule no deletion. A payment of S$48,000 to K. Ong waits for your approval, and you approve it. SEND report to finance is allowed. READ payment credentials is blocked by the rule no secrets. Each decision becomes a signed record, #001 to #005, linked to the one before.
Finance agent Checked by TraceStone
Example

Your agents act in seconds. Your controls see it later.

AI agents now move your money, change your records and use your tools. Most companies only find out what happened afterwards, from the logs.

Three real incidents

With TraceStone, nothing runs before it is checked.

The rules are written. Your deadline is close.

Regulators now expect you to control and record what your AI does. In Singapore, the MAS Guidelines on AI Risk Management set clear dates for financial institutions.

  • Show human oversight Prove that a person approved the risky actions.
  • Keep records of AI use Show what each agent did, and when.
  • Check before action Follow the MAS SAFR advice: check each action before it runs.
Learn more about the rules and your deadlines

Key AI dates, Oct 2026 to Dec 2028

  1. 2026
  2. Oct 2026 · MAS AI Guidelines , Published
  3. Oct 2026 · SAFR v1.1: check actions first , Published
  4. Today · Nov 2026
  5. 2027
  6. Oct 2027 · AI inventory expected , in 11 months
  7. Dec 2027 · EU AI Act high-risk rules , in 13 months
  8. 2028
  9. Oct 2028 · Human oversight expected , in 23 months
11 months until your first deadlineAI inventory expected · 7 Oct 2027

Sources: MAS Guidelines on AI Risk Management (Oct 2026) · MAS SAFR paper v1.1 · EU AI Act

Bring your rules. We do the checking.

TraceStone runs inside your company, between your AI agents and the tools they use.

Example
Your company Data stays inside
Your AI agent
TraceStone
Your tools

Your rules

  • No deletion
  • Reading: allowed
  • > S$10,000: a person

Your signed records

  • READ revenue allowed
  • DELETE customer_records blocked
  • PAY S$48,000 waiting you approved
proof only
Your auditor gets PAY S$••••• → ••••• Rules held
Your auditor

No sensitive data
reaches the auditor

Your rules. Enforced before every action.

Set your rules once. Every agent action meets them before it touches your systems.

  • Stop what breaks your rules A blocked action never runs.
  • Approve what is risky Payments over your limit wait for you.
  • Rules that fit each task Built from each request and verified before use.
  • A record of every decision Signed, so you can show it later.
Learn more about runtime control
Example

Example: an AI agent tries three actions. DELETE customer_records breaks the rule "no deletion", is blocked and becomes record #005. READ revenue is allowed and becomes record #006. PAY S$48,000 to K. Ong is over S$10,000, so it waits for the CFO; the CFO approves, the payment is sent and becomes record #007.

No one can quietly delete your records. Not even an admin.

Every agent action is recorded and sealed. If anyone removes or changes a record, you and your auditor see it at once.

  • Signed records Every decision is signed, so no one can fake it.
  • Linked records Each record seals the one before, so gaps show at once.
  • Independent checkpoints Signed checkpoints go to an independent service, so even we cannot rewrite your history.
Learn more about tamper-proof records

An admin deletes one record

Example
Ordinary logsPlain text, no seal
  1. READ totals ok
  2. SEND report ok
  3. DELETE records blocked
  4. READ revenue ok
No check Nobody notices
Your records in TraceStoneSigned and linked
  1. #003 READ totals prev e9514218 hash 7f7ac212
  2. #004 SEND report prev 7f7ac212 hash 42866251
  3. #005 DELETE Blocked prev 42866251 hash 39660d73
  4. #006 READ revenue prev 39660d73 hash eb124917
  5. Signed checkpointSent to an independent service last eb124917
Chain intact Tampering detected

Four agent actions, recorded in both places.

Pass your audit. Keep your data private.

Your auditor asks for proof that every rule was followed. With TraceStone, you hand over proof, not your customers' names and amounts.

  • Share only what is needed Your auditor sees only what the audit asks for.
  • Prove without revealing Zero-knowledge proofs confirm a rule held, without the values.
  • Sealed checks Complex checks run in a sealed area of the processor (TEE).
  • Your auditor checks alone An offline verifier, with no access to your systems.

Learn more about private audit

A hidden instruction: PAY S$48,000 → K. Ong

Example
AI agent
No checkSecurity toolTraceStone > S$10,000: a person
K. Ong

Payment

SENTto K. Ong

BLOCKEDby the security tool

Waiting for a person

BLOCKEDyour CFO said no

Record

14:02:11 PAY 48000 okPlain log, easy to change
Tool consolePAY 48000 · blockedLog in the tool's console
#007#008 signedSigned, linked record

Your auditor asks: did every payment over S$10,000 have approval?

What your auditor gets

Nothing to proveOnly a plain log to trust.

Full logs. Every name and amount is visible.

Yes. Every payment over S$10,000 had approval. Zero-knowledge proof

#101 J. Tan S$12,400 approved
#102 R. Lim S$3,100 under limit
#103 K. Ong S$48,000 blocked

+ 1,237 more payments

Logs ask your auditor to trust you. Proof lets them check.

Other tools give you logs. TraceStone gives you proof.

Cloud platforms protect agents inside their own platform. Security tools cover many platforms, but your auditor still reads raw logs. TraceStone covers all your agents and gives your auditor proof.

  • All your agents Not just one cloud or one product.
  • Proof, not logs Your auditor checks a proof instead of reading records.
  • Your data stays with you Nothing sensitive leaves your environment.

Learn more about how TraceStone compares

Criterion
TraceStone
CrowdStrike Falcon Guardian
AWS AgentCore Policy
ServiceNow AI Control Tower
What your auditor sees
Proof that the rules held, not the customer data
Raw records, with redaction options
Logs, protected by access control
Records inside the platform
Where your auditor checks
Anywhere, offline, with a standalone verifier
In CrowdStrike's console
In the customer's AWS account
Inside the ServiceNow instance
Tamper-evident records
Each record is signed and linked to the one before
Not stated (audit trail only)
Only for activity logs, not decision logs
Not stated (audit records only)
Task-aware rules
Rules derived from each user task, verified before use
Not stated
Not stated (fixed rules, checked with automated reasoning)
Not stated
Runtime protection and human approval
Both built in, for any agent framework
Blocks risky prompts and behaviour; approval only for its own agents
Checks and approval rules, but only through AWS's gateway
Controls which agents may be used; approval only for ServiceNow-built agents
Zero-knowledge proof for your auditor
Yes
Not stated
Not stated
Not stated
Where it runs
Inside your own environment
CrowdStrike cloud
AWS cloud
Runs on the ServiceNow platform

Also checked

  • Microsoft Foundry — guardrails cover Foundry agents only.
  • Google Agent Gateway — no approval step, no signed records.
  • Palo Alto Prisma AIRS — raw sessions stay in the console.
  • Cisco AI Defense — alerts on violations, not proof.
  • SentinelOne Prompt Security — no approval step.
  • Agent-security start-ups — no signed records an outsider can check.
Based on each company's public documentation, checked October 2026. Names belong to their owners.

Your environment. Your data.

Run TraceStone on your own servers, in your private cloud or in a hybrid setup. It works with the agent framework you already use.

Start with the log collector. Move to the gateway when you are ready.

Learn more about deployment

Your environment

  1. Security gateway

    Put TraceStone in front of your agents' tool calls. It can block.

  2. Embedded check

    Call TraceStone from your agent's own code. It can block too.

  3. Log collector

    Send records after the fact. Start by observing, with almost no changes.

Example

SECURE THE FULL AI LIFECYCLE

From security review
to runtime evidence.

Keep using the TraceStone products for application testing and agent security alongside runtime control and private audit.

BEFORE DEPLOYMENT · APPLICATION SECURITY

Vulnerability discovery

Find security gaps in AI gateways, agents and LLM applications with AI-focused testing, penetration testing and code review.

Explore vulnerability discovery

BEFORE DEPLOYMENT · AGENT SECURITY

Agent security

Review an agent’s code, tools and permission model to surface prompt injection paths, privilege risks and trust-boundary issues before release.

Explore agent security

AFTER DEPLOYMENT · MONITORING

Evidence-AIDR

Monitor deployed agents, detect attacks and seal runtime events into tamper-evident evidence that can be checked offline.

Explore Evidence-AIDR

See it on a real agent.

Watch a reporting agent at work: safe actions pass, an export waits for your CFO, and a hidden instruction is stopped. Every action is recorded.

TraceStone· Live requests Example

Good questions. Straight answers.

Still curious? Talk to us

What does TraceStone do for us?

It checks each of your agents' actions before it runs, asks a person before risky ones, keeps signed, linked records, and gives your auditor proof that the rules held, without your customers' data.

How do we stay in control?

You set the main policies. Task-aware rules come from each user request, the available tools and the workflow, and are checked against your policy and formally verified before use. High-risk actions wait for your approval.

How do we prove it to our auditor without giving them our data?

You hand over proof, not your data. You export only the records the audit needs, so your revenue figures, prompts and replies, and request payloads stay inside your company. Zero-knowledge proofs then show that each rule held without revealing the values, for example that an action had a valid approval without showing who approved it. Your auditor checks the proofs and the signed records with an offline verifier, so they never need access to your systems.

What does our auditor receive?

Only the evidence the audit needs. Zero-knowledge proofs show that a rule held without revealing the value, for example that a payment stayed below its limit. Complex checks can run in a trusted execution environment (TEE). An offline verifier marks each claim as verified, failed or inconclusive; missing evidence never counts as a pass.

Can anyone change our records?

Each record is signed and linked to the one before it. A Merkle tree lets your auditor check one record without the whole log, and signed checkpoints go to an independent service. A removed, added or changed record is detected.

How do we connect our agents?

Three ways: a security gateway in front of your agents' tool calls, an embedded check in your agent's code, or a log collector that records after the fact (it cannot block, but keeps signed records). You can start with the log collector and move to the gateway later.

Where does our data go?

Nowhere new. TraceStone runs inside your environment: your servers, your private cloud or a hybrid setup. It works with any agent framework.

Which rules does it help us meet?

The MAS Guidelines on AI Risk Management (human oversight and records of AI use), the MAS SAFR paper (check each action before it runs) and the EU AI Act (logs and human oversight for high-risk AI).

How do we start?

With a paid pilot of 6–8 weeks on one of your agent workflows, inside your environment.

Still curious? Talk to us