layerline

Manage Layerline project assets — create, update, archive, restore, comment, and track output completion. Use whenever the user asks Claude Code to add a new asset to a project, change an asset's status, browse assets, comment on an asset, mark a modeling/texturing/rigging output as done, or configure how an asset is exported. Covers the asset CRUD tools, asset-comment threading with @mentions and

Foundational context for using the Layerline MCP server. Use whenever the user asks Claude Code to do anything in Layerline — listing projects, finding orgs, resolving member IDs, getting dashboards, checking online users, or before running any other layerline:* skill. Covers the `mcp__layerline__*` tool namespace, org/project resolution rules, the API-key vs JWT distinction, common ID-lookup patterns, and the project status/role model. Invoke this skill BEFORE asking the user for an orgId, projectId, or userId — usually you can resolve them yourself.

Drive AI generation in Layerline (GenStudio) — concept images, image-to-3D, text-to-3D, textures, retopology, LOD, rigging, animation, and audio SFX. Use whenever the user asks Claude Code to generate an image, model, texture, rig, or animation; build a multi-step character pipeline; persist a generation as a project asset; expand a short idea into a polished prompt; manage style presets; or read a Game Design Document (GDD) for context. Covers the unified `generate` entrypoint (one-call generate-and-save), provider routing (auto strategy or explicit slug), the multi-step `generate_character_pipeline` (concept → model → texture → rig → animate), generation history and version chains, presigned download URLs for results, prompt enhancement, reusable style presets, GDD hydration for art-direction context, and moodboards as visual references.

Run the Layerline review/approval flow — create reviews, add files and reviewers, leave threaded comments with severity and type, submit per-reviewer or per-file decisions, attach checklists, and use AI helpers for change summaries and feedback rewriting. Use whenever the user asks Claude Code to start a review on an asset or task, request approvals, comment on a review, mark a comment resolved, approve/reject/request-changes, run a pre-submit check, or estimate approval readiness. Covers the full review lifecycle (draft → submitted → in_review → decided → reopened/superseded), reviewer roles (reviewer/approver/final_approver/observer), and AI-assisted tools that summarize comments, rewrite vague feedback, and score readiness.

Manage Layerline Storage (LS) files and the per-entity File Boxes that group them. Use whenever the user asks Claude Code to find a file in Layerline by path, list files in a project or folder, get a download URL, restore an old version, lock/unlock a file before editing, share a file or folder with an external link, configure cloud sync rules, or attach files to an asset/task/shot/GDD-asset's File Box. Covers file ops (list/find/rename/move/delete), version history and restore, exclusive locks with audit trail, presigned download URLs (~1h TTL), share links (public/password/email-invite), include/exclude sync glob rules, storage usage stats, and the File Box pattern for grouping LS files under a higher-level entity.