chore(release): prepare Grid v0.1.4

[nerdalert]

Summary

Prepares Grid v0.1.4 for release with updated package and Helm chart versions, release notes, official Praxis AI defaults, and reproducible qualification image references.

This release includes the provider-selection and distributed token-quota work merged through PRs #65 and #84.

Highlights

• Provider selection groups and round-robin request placement.
• Stable provider attribution across regional gateways.
• Distributed, Valkey-backed sliding-window token quotas.
• Atomic quota reservations across multiple consumer gateways.
• Natural quota expiry and recovery.
• Quota persistence across gateway restarts.
• Fail-closed behavior during Valkey outages.
• NetworkPolicy isolation for shared quota state.
• Pressure-aware provider admission and recovery.

Release Packaging

• Bumps the Grid workspace and Helm charts to 0.1.4.
• Changes the Praxis gateway chart default to the official ghcr.io/praxis-proxy/ai:0.3.0 image.
• Verifies the official AI image digest and OCI provenance during release.
• Updates qualification topologies to reference Grid v0.1.4 artifacts.
• Adds detailed v0.1.4 release notes.
• Removes the previous Grid AI rollup from the release workflow.

Praxis AI Image Requirement

The official Praxis AI 0.3.0 image supports provider selection and load balancing.

The distributed quota qualification additionally requires an AI image built with:

  token-rate-limit-filter,praxis-filter/basic-auth-filter

These optional filters are not included in the standard AI 0.3.0 image. The quota documentation provides the source-build procedure. Grid does not build or publish a replacement AI rollup.

Validation

The following checks passed:

• cargo +nightly-2026-03-28 fmt --all -- --check
• cargo clippy --workspace --all-targets -- -D warnings
• make test
• make doc
• make lint
• git diff --check
• Release workflow actionlint
• Helm lint and rendering for all four charts
• Forge validation for both qualification topologies

Runtime qualifications also passed:

• AI Intelligent Routing 0.3.0: 60/60 provider-routing requests succeeded, with 20 requests attributed to each of three providers.
• Feature-enabled AI build: shared quota, concurrent reservations, natural expiry, restart persistence, Valkey outage and recovery, and NetworkPolicy enforcement all passed.

[praxis-bot]

praxis-bot review: chore(release): prepare Grid v0.1.4

Purpose

Migrates Grid from a self-published AI rollup image to the official Praxis AI 0.3.0 image, bumps all workspace and Helm chart versions to 0.1.4, adds release process documentation, and updates qualification topologies to reference the new artifacts.

Assessment

Clean release preparation. All version bumps are consistent across Cargo.toml, Cargo.lock, and the four Helm charts. The release workflow correctly handles the divergent appVersion for praxis-gateway (tracks the Praxis AI version 0.3.0 rather than the Grid tag v0.1.4). The token-rate-limit topology correctly switches from the old rollup to a locally built praxis-ai:quota-features-required image with imagePullPolicy: Never, and the README documents the build procedure. The provider-traffic topology correctly references the official published image. The new docs/release.md is comprehensive and accurately describes the workflow.

The AI image verification step now pulls the image by immutable digest and inspects OCI labels for source, revision, and version provenance -- a stronger verification than the previous dual-tag digest comparison.

Findings

No Critical, Large, or Medium issues found.

Severity	Count
Critical	0
Large	0
Medium	0