roost-xr
claudecodexMeta Quest, Unity, and OpenXR development
19 skills · 7 agents
Mesh experiment agents and diagnostics
Plugin roost-mesh from marketplace the-roost in wesduffdev/the_roost. Commands read from the repo's manifest — 1 month ago.
mesh-experiment-reviewReview geometry and mesh experiments with reproducible evidence.
architectUse for architecturally significant decisions - new services, module boundaries, data flows, technology selection, cloud infrastructure, and cost analysis. Produces C4 diagrams in Mermaid, ADRs, and Terraform. Invoke when a change crosses a module boundary, introduces a dependency, alters a data flo
carrier-refuterUse to attack a claim that the carrier is lossless, exact, or has a stated capacity. Recomputes the arithmetic independently, hammers the codec with adversarial inputs, and reports what survives. Strictly read-only against tracked files. Invoke before any capacity or byte-exactness claim reaches a R
code-reviewerUse to review a branch, PR, or diff before merge. Runs the test suites, enforces the team's engineering principles, and judges design soundness. Strictly read-only - returns findings as JSON to the orchestrator and never edits code. Invoke after the developer reports a ticket complete and before any
developerUse to implement a ticket. Verifies the Definition of Ready before writing code, then works strictly test-first (red-green-refactor) and closes with BDD scenarios proving the acceptance criteria. Invoke for features, bug fixes, and refactors scoped to a Linear ticket.
experiment-developerUse to implement one slice of a validation experiment that is specified by a design document rather than a Linear ticket. Works strictly test-first (red-green-refactor) against a named DESIGN.md section or approved plan file, and reports the real command output. Invoke for experiment work in experim
mesh-visualizerUse to put the rendered object in front of a human. Drives the viewer headlessly in Docker, captures screenshots of the mesh from named angles and shading modes, and measures load and render time. Owns scripts/sweep-vault-ladder.mts, the payload-ladder geometry sweep. Produces images and numbers; ne
orchestratorUse as the entry point for any multi-step or multi-agent request. Decomposes work, routes each piece to the right specialist, fans out parallel instances where the work is independent, and consumes the code-reviewer's JSON to dispatch fixes. Does not implement, review, or document anything itself.
release-managerUse for any git operation touching a shared ref - push, PR creation, merge to develop or main, release branches, tags, and hotfix back-merges. Enforces the Git Flow topology and the merge gate. Does not author or edit code, and does not resolve conflicts. Invoke after the code-reviewer approves.
researcherUse to understand and explain the codebase as it actually is - orientation for newcomers, tracing how a request flows, mapping a module's real interface, or assessing blast radius before a change. Produces cited explanations and Mermaid diagrams. Read-only; returns its report rather than writing fil
ticket-managerUse to create, refine, split, or update Linear issues. Selects the right template (story / bug), enforces testable Given-When-Then acceptance criteria, maintains the project → story hierarchy, ticks acceptance criteria and tasks on evidence, and posts confirmed comments. Invoke when work needs a tic
ticket-writerUse to turn a completed Definition-of-Ready interview into a ticket write-up in the ticket-manager templates, or to draft a plain-language comment for an existing issue. Applies the AC quality tests, computes the estimate band from the surface count, places the researcher's suggested tasks and Tech
wiki-writerUse to create and maintain GitHub wiki documentation. Researches the codebase, keeps pages current with the shipped code, and publishes to the wiki repo. Escalates to architect when documentation reveals an undocumented or incoherent design. Invoke when a public interface changes, an ADR lands, or d
Meta Quest, Unity, and OpenXR development
19 skills · 7 agents
Code review, release, and delivery practices
4 skills · 10 agents
React interfaces and dashboards
1 skill · 3 agents
Server architecture, APIs, and operations
1 skill · 3 agents
Opt-in hook guidance and local evidence gates
1 skill · hooks
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
Scaffold, validate, and publish new Claude Code plugins with proper structure
3 skills · 1 command · 1 agent
Up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your coding agent's context.
1 MCP server
Use Alibaba Open Code Review from local Codex.
2 skills
Claude Code skill for creating and presenting Slidev slidedecks
1 skill
Code intelligence powered by a knowledge graph. Provides execution flow tracing, blast radius analysis, and augmented search across your codebase.
12 skills · 1 MCP server · hooks
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.