# Meshery v1.0.8

2026-04-15 · [Original release notes](https://github.com/meshery/meshery/releases/tag/v1.0.8) · [Full summary](https://ratatosk.io/en/releases/meshery/v1.0.8)

Breaking changes 3 · Important fixes 1 · Enhancements 1 · Dependency updates 7 · Other changes 9

## Breaking changes (3)

### Dashboard widgets enforce default minimum dimensions

- Affects: Meshery UI dashboards with widgets

Before this release, users could shrink widgets below their default dimensions in edit mode, and older saved layouts did not enforce those limits. After upgrading, older saved layouts also enforce the minimum dimensions.

Evidence: “- Ensured older saved layouts from the database automatically respect these updated constraints.” https://github.com/meshery/meshery/pull/18673

> - feat(dashboard): enforce minimum size constraints for widgets @Rajesh-Nagarajan-11 (#18673)

### Concurrent relationship evaluations now share results

- Affects: relationship-policy evaluation in Meshery Server

When requests evaluate the same design concurrently, followers receive the leader's result, and only the leader's request persists the evaluation event.

Evidence: “// Coalesce concurrent evaluations of the same design (rage-click guard).” https://github.com/meshery/meshery/pull/17330

> - [Server] Policies: Port OPA/Rego policy engine to native Go @guan404ming (#17330)

### Relationship policy evaluations now have a three-minute timeout

- Affects: Meshery Server setups using the relationship policy evaluation handler

The timeout defaults to three minutes and is configurable with `POLICY_EVAL_TIMEOUT`. When an evaluation reaches the deadline, the handler returns HTTP 504, but its goroutine may continue running in the background.

> - [Server] Policies: Port OPA/Rego policy engine to native Go @guan404ming (#17330)
> - [Server] Policies: Port OPA/Rego policy engine to native Go @guan404ming (#17330)

## Important fixes (1)

### Relationship-evaluation response actions deduplicated

- Affects: relationship-policy evaluations with re-evaluation iterations (Inferred)

The accumulated response actions are deduplicated by `(op, id)` after the re-evaluation iterations.

> - [Server] Policies: Port OPA/Rego policy engine to native Go @guan404ming (#17330)

## Enhancements (1)

- Native Go engine for relationship policy evaluation (To use: Set `USE_GO_POLICY_ENGINE` to true to select the Go engine; OPA/Rego remains the default.)

## Dependency updates (7)

7 more, none mentioning security

- `- Bump follow-redirects from 1.15.11 to 1.16.0 in /install/docker-extension/ui @[dependabot[bot]](https://github.com/apps/dependabot) (#18710)`
- `- Upgrade meshsync and k8s deps to v0.35.2 @leecalcote (#18685)`
- `- [Chore]: Update to Sistent v0.18.7 @l5io (#18711)`
- `- Bump follow-redirects from 1.15.11 to 1.16.0 in /ui @[dependabot[bot]](https://github.com/apps/dependabot) (#18709)`
- `- [Chore]: Update to Sistent v0.18.6 @l5io (#18706)`
- `- Bump axios from 1.13.6 to 1.15.0 in /provider-ui @[dependabot[bot]](https://github.com/apps/dependabot) (#18628)`
- `- Bump helm.sh/helm/v3 from 3.19.4 to 3.20.2 @[dependabot[bot]](https://github.com/apps/dependabot) (#18650)`

## Other changes (9)

- Unit tests added for `EncodeRefUrl` and `DecodeRefURL`
- UI restructuring added to the v1.1.0 roadmap
- `mesheryctl` component view end-to-end test assertions updated
- Back tooltip positioned beside its button
- UI theme re-export and import alias added
- UI restructure lint guardrails and audit scripts added
- PR-preview workflow adds or updates a comment when previews are pruned
- PR-preview workflow retains the six most recently updated previews
- `mesheryctl` component-view assertions aligned with current CLI output

---

Summaries are produced by AI, so they can be wrong, and they are provided without warranty.
