For Risk, compliance and audit

When I ask who approved this, can anyone show me the full trail?

The answer is usually spread across several systems and someone's memory. LeafCraft keeps every task and decision, with the reasons of AI and people, on your side and open to your auditors.

One teamYour rulesYour memoryYour budget
Product screen · example dataLeafCraft console

What's changing for you

01

Logs are not records

A log says what happened. You need how it decided, who checked, and why.

02

The vendor holds the evidence

The trail your auditor asks for sits in someone else's stack, for a few months.

03

Agents change faster than policies

A model update inside an agent is a change nobody told you about.

Two risks the layer removes

The agents you can't see, and the model that gets retired.

Agents you don't know about

In your company, agents arrive through vendor updates and team pilots. LeafCraft finds all of them, gives each an owner, and brings it under your rules or removes it.

How discovery works

A model that gets retired

An agent built straight on one model has to be rebuilt and re-approved when that model is retired or repriced. On LeafCraft you switch the model for that agent in config and keep everything else. Months and a budget, saved.

Agent tied to one model
  • The model is retired or repriced
  • Prompts, tests and sign-offs rebuilt
  • Months of work, a new approval cycle
  • The history stays with the old tool
MonthsA rebuild budget
Agent on LeafCraft
  • Pick another model for that agent
  • Switching is a config change
  • Test it against your own record first
  • Rules, limits, owner and memory stay
HoursNo rebuild
TimeHours to switch, not months to rebuild
CostNo rebuild budget, no new approval cycle
MemoryThe record carries over to the new model
Your choice of model

What keeps you up

1

A log is not a record

A log shows what happened. A record shows how it was decided, by whom and why, including who approved and whether it was escalated.

2

The counterparty holds the evidence

When the vendor that sells the agent also keeps the record, the evidence sits with the party you are checking.

3

Maker and checker blur

With agents in the flow, it is easy for the same teammate to do the work and sign it off.

Why it matters now

One layer under every agent and every person.

Your question is simple and nobody can answer it: who approved this, and can you show me the full trail? The layer answers it for every decision by every agent and every person: the rule applied, the limit, the checker, both reasons, and every change to authority, kept unchanged in your environment and exportable in open formats.

How the layer works

What LeafCraft gives you

01

Both reasons, side by side

Every decision keeps the agent's reasoning and the person's reasoning together. Approvals, rejections and escalations are all on it.

02

A record you hold

Run records sit in your own Redis or Postgres, kept as long as you set. Rules and records export in open formats on demand.

03

A log that cannot be changed

Ready-made templates end every flow with an audit agent that writes to a log that cannot be changed once written.

04

Privacy built in

Message content can be masked in traces. Personal details are masked before an agent saves a lesson, and erase requests are recorded too.

What changes for you

Today
With LeafCraft
A log of what the agent did
A record of how it decided and who checked
Evidence held by the vendor
Evidence in your database, unchanged, exportable
Model changes you hear about later
Every model switch recorded and tested on your history
Maker and checker the same agent
Never the same, enforced on every run
Silence is never a yesEstimated before it runsNothing erased after it's written
Product screen · example dataLeafCraft console

Ask any AI agent vendor

Five questions for your next vendor meeting. They are the ones your auditor and your board will ask you. Ask them first.

  1. Who approved this decision, and was it escalated?
  2. Can you walk me through every step, then replay it?
  3. Do you keep the person's reasoning as well as the agent's?
  4. Where is the record held, for how long, and can it be changed after it is written?
  5. Can I export policies and records in open formats, on demand, without raising a project?

Questions teams ask

Is this a record of people's behaviour?

No. It is a record of the work: tasks, decisions, handoffs and the reasons behind them. It is not surveillance of people.

Can each agent reach only what its role needs?

Yes. Every agent has its own login with its role's access. A tool that is not on its list is refused before it runs.

What happens when work breaks a rule?

Work that misses your rules stays put until it is fixed, and it goes to the people you named. All of it is recorded.

Bring your hardest question.

We'll answer it on one of your own processes, with the record to show for it.