Vanta
Read a Vanta compliance program — failing tests and the resources causing them, controls, frameworks, issues, vulnerabilities against their SLAs, people's overdue tasks and the vendor inventory.
About
Vanta ships in the w6w first-party pack. It declares 25 actions, 3 health checks, and the host runs its code in a sandbox that never sees the credential.
- App id
io.w6w.vanta- Version
- 0.1.1
- Author
- w6w
- Licence
- MIT
- Categories
- Legal & Compliance · Security & Identity · Monitoring & Observability
Overview
Vanta brings a compliance program’s live state into a workflow — which tests are failing and why, which controls have an owner, which issues and vulnerabilities are approaching their deadlines, and which vendors and people are covered. Test actions read what is currently failing and which specific resources are causing it, with an action to formally exclude a resource from a test as a documented compliance exception rather than silently ignoring it.
Control, framework and policy actions read what an organization is held to and who owns meeting each requirement, while issue and vulnerability actions surface the work queue sorted by due date or SLA deadline, plus what has already been remediated and whether it happened in time. People actions track who the compliance program covers and their overdue tasks, with a dedicated offboarding action for completing that checklist when someone leaves. Vendor, integration and monitored-device actions round out the picture — including the accounts feeding evidence into Vanta at all, since a disconnected integration doesn’t fail its tests, it just lets them go stale.
This app deliberately never deletes compliance records, uploads evidence files, or acknowledges a missed security deadline on a workflow’s behalf — those stay decisions for a person to make directly in Vanta.
Build with Vanta
Three routes to the same 25 actions. The Workflow tab is generated from Vanta's own manifest and carries its real ids, so it is copy-pasteable; the Code and CLI examples are the same call for any action on any app, so every app-specific value in them is a blank you fill in.
Get a control
control-get One requirement with the tests and documents Vanta treats as its evidence — which is the question an auditor actually asks.
List controls
control-list The requirements a framework imposes. A control with no owner is a requirement nobody is accountable for — the most common finding in a readiness review, so they are listed apart.
Set a control's owner
control-set-owner Make a Vanta USER accountable for a control — not a person. The two rosters differ, and passing a person id from `person-list` will not work.
List documents
document-list The evidence that is not automated — reports, minutes, signed agreements. A document Vanta expects and has never received is IN this list with a status, not missing from it.
List event logs
event-log-list Vanta's own audit trail — who granted an exception or reassigned a control. Reads forward from a start date, which is the shape a tailing job wants.
List a framework's controls
framework-control-list One framework's requirements in its own structure — how a readiness report is organised. The same control appears under several frameworks with different numbering.
List frameworks
framework-list The compliance frameworks this tenant tracks. Most other filters take a framework id, and the same test means different things depending on which certification it evidences.
List connected integrations
integration-list What feeds Vanta its evidence. A disconnected integration does not fail its tests — they go stale and stay green, which is why this is worth checking before trusting a report.
Get an issue
issue-get One issue, with the readable id a person quotes and the controls it maps to — a ticket naming the requirement it threatens gets prioritised differently.
List issues
issue-list Vanta's work queue — items with an owner, a due date and a lifecycle. Filtering to a due date next week produces a list somebody can still act on.
List monitored computers
monitored-computer-list Laptops and their compliance state. A machine that has stopped reporting has no state at all — worse than a failure, and invisible in a report that counts only failures.
List people
person-list Everybody the compliance program covers — NOT the same as Vanta users. Filter by task status to find outstanding training, policy acceptances and background checks.
Offboard people
person-offboard Complete Vanta's offboarding checklist for people who have left. The acknowledger is recorded in the audit trail — it should be somebody who can be asked about it.
List policies
policy-list Written policies and their approval state. A policy updated last month resets every acceptance, which is why the reminder workflow is a recurring job rather than a one-off.
List risk scenarios
risk-scenario-list The risk register — the part written by people rather than collected by integrations. Enterprise risks are excluded unless asked for, which is why a register can look short.
Deactivate a test entity
test-entity-deactivate Exclude one resource from a test. This does not fix anything — the test goes green and the condition stays. The reason goes in the audit trail, and an exception without an expiry outlives its justification.
List the entities failing a test
test-entity-list The actual resources causing a test to fail — the buckets, repositories, laptops or accounts. Without this a workflow can only forward a red light to a human.
Get a test
test-get One test with the controls it satisfies and the integration that feeds it — which is what distinguishes 'this is broken' from 'the integration is broken'.
List tests
test-list Continuously-evaluated compliance assertions and their status. `NEEDS_ATTENTION` is the one that means broken — `IN_PROGRESS` and `DEACTIVATED` are not failures.
List Vanta users
user-list People with a Vanta login — not the same roster as `person-list`. Every ownership field in the API takes an id from here.
Get a vendor
vendor-get One supplier with its risk level, last review and the data it handles — which is what separates an inventory from a risk picture.
List vendors
vendor-list The third-party inventory frameworks require. Filter by status to get the review queue — listing every vendor answers nothing.
Get a vulnerability
vulnerability-get One finding with the package, the fixed version if there is one, and the CVE. A finding with no fixed version is monitoring; one with a fix is an upgrade.
List vulnerabilities
vulnerability-list Scanner findings with their remediation deadlines. Filtering to a deadline a week out is the query worth scheduling — running it against now lists failures instead.
List vulnerability remediations
vulnerability-remediation-list What was fixed and whether it met its deadline — the record an auditor asks for, since the requirement is not 'have none' but 'remediate within the window'.
A workflow step names the app and the action, and the editor fills in the
connection when you pick one. This is the Step shape from the
workflow spec, carrying Vanta's real ids.
{
"manifestVersion": "2",
"name": "vanta-example",
"steps": [
{
"id": "control-get",
"uses": {
"app": "io.w6w.vanta",
"action": "control-get",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"controlId": "<controlId>"
}
}
]
}Here are some of the things you can do
-
Get a control
readcontrol-get -
List controls
readcontrol-list -
Set a control's owner
performcontrol-set-owner -
List documents
readdocument-list -
List event logs
readevent-log-list
+20 more actions available
Every app-specific value here is a blank you have to fill in. An
app action is reached through the connection that authenticates it, so the
address is a connection id, not the app id — and connections belong to your account,
so a public page cannot know yours. Create one for Vanta, then fill in
the three blanks: conn_YOUR_CONNECTION_ID, the action key, and the
parameters that action declares. The call itself is real — the shape is transcribed
from the studio's own snippet builder, which prints the same kind of blanks — but
nothing in it is specific to Vanta. The Workflow tab is where this app's
real ids are.
npm install @w6w/sdkyarn add @w6w/sdkpnpm add @w6w/sdkdeno add npm:@w6w/sdkimport { W6wClient, isActionRun } from "@w6w/sdk";
// Reads W6W_BASE_URL and W6W_TOKEN from the environment when omitted.
const client = new W6wClient();
const envelope = await client.run({
urn: "conn_YOUR_CONNECTION_ID",
action: "control-get",
payload: {
controlId: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value);npm install -g @w6w/cliw6w run conn_YOUR_CONNECTION_ID --action control-get --payload '{"controlId":"<value>"}'Give an AI agent Vanta — without giving it Vanta's credentials. One MCP endpoint exposes every app, function and workflow the caller is entitled to, as tools it can discover and run. Access is granted per team while we onboard.
One tool call{
"name": "w6w_invoke",
"arguments": {
"ref": "app:io.w6w.vanta#control-get",
"input": {
"controlId": "<controlId>"
}
}
}
Every tool names its target with a single ref. The
app: form above doesn't name a connection at all — the
host resolves which of the caller's Vanta connections to sign
with, and refuses rather than guesses when the answer is ambiguous.
What the agent gets
Credentials it can't read
The token is attached host-side, at the moment of the call. It is never a tool argument, never in the model's context, and never in a transcript — so a prompt injection has nothing to exfiltrate.
A tool surface scoped to the caller
Tools are derived per end user from what that person has actually connected and is entitled to — not one shared bot identity carrying the union of everyone's access.
A durable workflow in one call
Multi-step work runs on the workflow engine and returns a run handle the agent can poll — retries, branching and state survive the conversation that started them.
Health-aware discovery
Vanta's declared health checks are on the surface too, so an agent can tell "the vendor is down" from "your credential expired" before it burns a retry on either.
The MCP surface is part of the hosted platform. Vanta itself is MIT, and the runtime that executes it is source-available (FSL).
Health checks
Vanta declares its own checks, so its health is a property of the app rather than something the host guesses at.
Vanta platform status
The public API this app calls, plus the integrations that feed it. An integrations outage is capped at degraded — the API keeps answering, and its answers go stale.
Tenant reachability
Whether this connection's own Vanta tenant answers. A 401 here usually means another process minted a token for the same application, not that the secret is wrong.