# What MeshQu does

Canonical: https://www.meshqu.com/knowledge/what-meshqu-does

What does MeshQu actually do?

MeshQu assesses consequential decisions against the applicable policy and keeps a signed record of each recorded assessment. Your systems control the action: MeshQu returns an advisory verdict. The rules come from the approved policy; the receipt records that policy version, the supplied context and the outcome. Bring your own agent through MeshQu’s OAuth-secured MCP server or documented API. Availability is for research, evaluation and design-partner deployments.

## In more detail

Start with one consequential decision and the policy behind it. The homepage illustrates supplier onboarding: a missing tax registration certificate sends the supplier for review, a later assessment can be linked to the first, and the final call and its rationale can be recorded. This is illustrative example data, not a customer deployment. In the sidecar shape your workflow continues to control the action while MeshQu assesses and records alongside it. In the in-path shape your application asks for a policy verdict before an action, then proceeds, requests review or stops. Evaluation alone does not persist a decision; recording is the operation that keeps the decision and receipt. Evidence content remains with its custodian: MeshQu binds supplied references and digests. Actor identifiers are supplied by your systems and are not authenticated by MeshQu. Qu, the assistant on this website, explains MeshQu and can help you scope a 30-minute mapping call. It cannot see your organisation’s records. Questions about recorded decisions are illustrative and depend on what a deployment records and makes available. The Solutions pages illustrate AI agent governance, human-in-the-loop automation and enhanced due diligence; these are applications, not claims of deployed sector coverage.

## Limits

- Your systems control the action. MeshQu returns an advisory verdict and does not independently enforce it.
- The record covers assessments recorded through MeshQu, not every decision an organisation makes; evaluation alone does not persist a decision.
- The rules come from the approved policy. The receipt description here is the policy, rules checked, decision context and outcome; it makes no claim about the outcome for each individual rule or the evidence for each rule in the receipt.
- Evidence references and digests are not custody of the evidence content, and actor assertions are not authenticated identity.
- Record integrity does not prove true inputs, a correct decision or that an action was carried out.
- The Northgate story and Qu questions about recorded decisions are illustrative. No public customer deployment, operational track record or working website-to-decision-records integration is claimed.
- Bringing your own agent through the MCP server or the API is offered within research, evaluation and design-partner availability. No general availability or named vendor integration is claimed.
- MeshQu is currently available for research, evaluation and design-partner deployments. Production availability and support commitments are agreed per engagement.

## Where it applies

- Map a decision that is made often, has costly errors, draws on evidence from several places, and may need an explanation later.
- Use supplier onboarding as one illustrative question; trade and credit, payments, insurance and AI-agent actions are other candidate decisions to assess with the team, not claims of deployed sector coverage.
- Answer a reviewer from the recorded policy, context, outcome, actor assertions and evidence references behind an assessment.
- Inspect patterns only in the decisions and fields a deployment actually records and exposes; illustrative Qu questions do not establish a live records integration.

## Sources

- [MeshQu docs — Concepts overview](https://docs.meshqu.com/concepts/overview): Decision evaluation versus recording, the policy context and the advisory verdict.
- [MeshQu docs — Integration patterns](https://docs.meshqu.com/guides/integration-patterns): The two documented shapes: assurance beside the decision, and a policy check in the path before an action.
- [MeshQu docs — Decision chains](https://docs.meshqu.com/guides/decision-chains): How linked receipts connect the assessments, human reviews and final call behind one decision.
- [MeshQu docs — Responsible AI: Evidence and Intent](https://docs.meshqu.com/concepts/responsible-ai-evidence): The current public availability sentence and the official Decision Receipt definition with its evidential boundaries.

## Related answers

- [Decision Receipts: what the record contains](https://www.meshqu.com/knowledge/decision-receipts): What is in a Decision Receipt, and what does it prove?
- [The limits of verification](https://www.meshqu.com/knowledge/limits-of-verification): What does verifying a Decision Receipt actually prove?
- [What a policy is, and how it is applied](https://www.meshqu.com/knowledge/policy-application): What is a policy in MeshQu, and how is it applied to a decision?
- [Independent verification: the bundle, and what replay checks](https://www.meshqu.com/knowledge/independent-verification): How does someone independently verify a Decision Receipt, and what does replay check?
- [What MeshQu is not](https://www.meshqu.com/knowledge/what-meshqu-is-not): What is MeshQu not, and which problems does it deliberately not solve?
- [What a receipt evidences about a responsible-AI commitment](https://www.meshqu.com/knowledge/responsible-ai-evidence): Can a Decision Receipt show that an AI decision was fair, accurate or correct?
- [What a receipt records about who acted](https://www.meshqu.com/knowledge/actor-identity): Does a Decision Receipt prove who took the action?
- [Evidence, references and source custody](https://www.meshqu.com/knowledge/evidence-and-source-custody): What does MeshQu do with the evidence behind a decision, and does it establish that the evidence is true?
- [AI agent governance, and bringing your own agent](https://www.meshqu.com/knowledge/ai-decision-accountability): How does MeshQu govern decisions an AI agent proposes, and can we use our own agent with it?
- [Human-in-the-loop automation with MeshQu](https://www.meshqu.com/knowledge/human-in-the-loop-automation): How can we automate checks in a review process while people keep authority over the decision?
- [Enhanced due diligence decisions with MeshQu](https://www.meshqu.com/knowledge/enhanced-due-diligence): How does MeshQu support an enhanced due diligence (EDD) decision, and what does it not do?
- [Decision chains: following a decision through reviews and the final call](https://www.meshqu.com/knowledge/decision-chains): How does MeshQu show a decision that took several steps, reviews and an exception?

## Next step

[Map a decision](https://www.meshqu.com/#model-a-decision)

Read more on [meshqu.com/receipts](https://www.meshqu.com/receipts)

Updated 9 October 2026
Machine-readable: https://www.meshqu.com/knowledge/what-meshqu-does.json
All reviewed answers: https://www.meshqu.com/knowledge