App · Reviews

Multi-round approvals with AI helpers.

A real review workflow — draft, submitted, in-review, decided, reopened, superseded. Per-file approve / changes / reject. Threaded comments with severity. Annotation overlays. Plus three AI helpers that summarise changes, summarise comments, and estimate approval readiness.

Get started for free

See the features

reviews.layerline.co

Screenshot · Review Studio

The problem

What this app solves.

Approvals happen in email and Slack threads that no one can find later. Frame.io is too narrow for a full creative pipeline. Half the time, "approved" means "the latest message was a thumbs-up emoji."

01 No clear handoff between artist and approver
02 Comments scattered across DMs and screenshots
03 Which version was approved? — unanswerable a week later
04 Approvers have no idea what changed since the last round

Features

Built for the way teams actually work.

01 · feature

Multi-round workflow

Reviews have explicit states — draft, submitted, in-review, decided, reopened, superseded. Every transition is logged. There's always one answer to "where is this in the process?"

create_review · submit_for_review · change_review_status · reopen_review

02 · feature

Per-file decisions

One review can hold many files. Each one gets its own decision — approve, request changes, or reject. Aggregate into the review's decision automatically.

submit_file_decision

03 · feature

Threaded comments with severity

Comments are threaded and severity-tagged — normal, critical, minor. Resolve or reopen each one. The thread is the audit log.

create_review_comment · resolve_review_comment · reopen_review_comment

04 · feature

Annotation layer for visual feedback

Draw on the file preview. Markup overlays sit on top of the asset so feedback points at exactly what the reviewer means.

components/reviews/ReviewAnnotationLayer.tsx

05 · feature

Three AI helpers

Summarise what changed between rounds. Summarise the comment pile. Estimate approval readiness. Each is an MCP tool — usable inside the product or from your own agent.

generate_review_change_summary · summarize_review_comments · estimate_approval_readiness

Use it from your code

Native MCP. Full REST.

Every endpoint Layerline uses internally is yours. Call it from an agent, from a CI step, from your DCC plugin. No vendor SDK, no rate-limit surprises.

review.example.ts

// Submit a per-file decision
await client.callTool({
  name: "mcp__layerline__submit_file_decision",
  arguments: {
    reviewId: "rev_44ab",
    fileId: "f_8s3m1",
    decision: "changes_requested",
    note: "Spec map reads as plastic in the third pass.",
  },
});

// Ask the AI what changed since last round
const summary = await client.callTool({
  name: "mcp__layerline__generate_review_change_summary",
  arguments: { reviewId: "rev_44ab" },
});

Plays well with

The rest of the platform.

App · Assets

One OS. Every app. Yours.

Start free. No credit card, no procurement loop. Scale to a studio when you're ready.

Get started for free Why Layerline