New: Introducing the Lemma MCP Server, and what we learned building it.New: Introducing the Lemma MCP Server. Read the blog post

Skip to main content

Grant Access. Bound Agents.Trace Every Call.

Governance is part of how Lemma runs work, not a layer added after. Each person, team and service gets the actions it needs on each resource, each agent gets only the tools its task requires, credentials and sensitive values stay protected, a person decides when a guardrail trips, and every request traces back to an identity.

Governance Built Into Every Run.

Before a Worker or an agent touches a consequential system, someone has to say who can act, what the agent can reach, how secrets stay secret, when a person decides, and how to prove what happened. Here is how Lemma answers each, followed through one payroll exception.

Access

Control exactly who can act on each resource.

Grant roles on the tenant or on a single resource. Payroll Operations can run the Payroll Adjustment Worker, Engineering can edit it, and its admin can share or delete it. A service account can be granted that one Worker, so an integration runs it without broad administration rights. Every request is checked against those grants, and a refused change is audited.

Roles per resource
Admin, Editor, Invoker and Viewer on a Worker; Admin, Editor and Viewer on a credential.
Service accounts
An identity for an integration or a scheduled job (a service principal), with its own credential that expires or is revoked. Its secret is shown once.
SSO and SCIM
Sign in through your OIDC provider; Okta or Microsoft Entra ID provisions, deactivates and groups users.
Access review
Each Worker and credential lists who has access, and with which role, in its access settings.
Tenants and regions
Workers, runs, functions and credentials belong to one tenant, an isolated workspace pinned to one region.
Agents

Give agents only the tools and authority their task requires.

An agent starts with no tools. Its author adds each one, a function or an MCP tool, with the credential it calls with. A parameter can be pinned to a trusted value from the run: the payroll agent searches policy freely but looks up only the employee on the case, because the Worker supplies that ID and the model never sees the field. The payroll write is not on the list at all.

  • No tools until an author adds them
  • Each tool calls with the credential its author chose
  • Pinned parameters come from the run, not the model
  • Max iterations and output tokens bound the loop
  • Every tool call shows in the run’s logs
Credentials and data

Protect credentials and sensitive data throughout a workflow.

A Worker names a credential instead of holding it, and stored credential data is encrypted with a key for its tenant. Each API call chooses whose access it uses: one service credential, each user’s own, or, where enabled, one chosen from the run’s data. Mark an input sensitive and protected views of the run mask it, and where enabled, so do supported copies and outputs made from it. Execution still uses the real value, so the calls and files a Worker sends it to remain the author’s responsibility.

By name
Workers name a credential; the secret stays in the store
A key per tenant
Stored credential data encrypted with its tenant’s key
Lemma Credentials
Managed credentials, enabled per tenant for supported functions
Integration data
Data access granted per Worker, per run and per operation, for a set time
Guardrails

Give a human control when a guardrail trips.

Put the check in the Worker, not in the prompt. The agent proposes a payroll adjustment, a rule compares it with policy, and anything over $5,000 goes to a payroll review that waits for approve, correct or decline. The payroll function runs only after an approval, and the same run continues with the reviewer’s answer. You set the condition, who is notified, the paths and what a timeout does.

The model proposes
Its output is an input to the rule, never the decision.
The rule routes
A rule checks the amount the same way on every run.
Values withheld from review
With the sensitive-data policy active, supported sensitive values stay out of the review; a field left blank keeps the value already supplied.
Step-up, where enabled
When a supported API asks for stronger sign-in, the run pauses and sends a one-time link to sign in again with your identity provider.
Audit

Trace who acted, under which identity, and what happened.

Security audit events record each covered change: the tenant, the identity that acted and the one it acted as, the request, its status and a trace ID. Refused changes are recorded too. The run’s logs answer the other question: which steps and functions ran. Where enabled, a user permitted to act as a service account can run a Worker as that account, and the audit names both.

  • Acting and effective identity
  • A trace ID on every event
  • Refused changes audited
  • Run logs per step
  • Immutable published versions
  • Run visibility by permission

Hand Agents Real Work.Keep the Authority.

The model proposes, the Worker applies the rule, a person decides the exception, and only an authorized function writes. Each part runs under an identity you granted, and every request lands on the record.

Walk Through It With Your Security Team.

See grants, agent tools, a guardrail and the audit trail with our team.

Request a Demo