the signals and their derived score are provisional and subject to change.
validate manually before sharing.
Total
0 /100
BAD
F
Total verdict of all categories combined. A score of 100 is best.
Likely LLM-assisted/authored commits found; total hits, and % of all commits.
LLM-related files found in the repository worktree.
LLM-related files found named in '.ignore'-files.
86c4025a5356e7ca8e505ee772254bc517903c8319cac80b303a1 month agoCo-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>d9da00ba6b1f2 months agoCo-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>7aaa5adaad212 months agoCo-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com># Levyra Claude Code
@AGENTS.md
Claude Code reads this file natively at project startup. The imported
`AGENTS.md` is the cross-runtime Levyra contract and must be applied before
repository work.
Claude-specific settings, hooks, agents, and path rules live under
`.agents/claude/` and may be projected locally to `.claude/`. Those runtime
helpers strengthen this contract but are not required for this root bootstrap to
work.
Do not duplicate the imported engineering rules here. Keep this bridge short so
sta# Levyra Claude Code Runtime
Root `CLAUDE.md` is Claude Code's native startup bridge and imports root
`AGENTS.md`, the cross-runtime source of truth. This file adds only
Claude-specific runtime behavior. Do not duplicate the full repository contract
here and do not preload large `docs/ai/` playbooks wholesale.
The generated `.claude/` projection may load this file, hooks, rules, agents, and
skills after `scripts/sync_agent_runtime.py` runs. The tracked root bridge means
Claude still receives the essential# Levyra Desktop Module Instructions
These instructions extend the root `AGENTS.md` for every file under `desktop/`.
## Module scope
Levyra Desktop is a Windows client built with Kotlin/JVM and Compose Multiplatform. It shares the extractor and localization catalog with Android and plays audio through libvlc.
Before editing, read `desktop/README.md` and load `.agents/skills/levyra-desktop/SKILL.md` plus any additional matching security, extractor, CI, review, or release skill.
For any visual redesign, # Levyra Engineering Instructions
## Purpose
This is Levyra's compact always-loaded engineering contract. Keep permanent
rules here and load detailed procedures only for the current task phase.
Instruction order:
1. root `AGENTS.md`;
2. nearest path-specific `AGENTS.md`;
3. approved `docs/project/` requirements when relevant;
4. at most the routed skill bodies needed for the current phase;
5. current code, tests, build files, workflows, and runtime evidence;
6. runtime-specific configuration under `.agen# Levyra Documentation Instructions
These instructions extend the root `AGENTS.md` for documentation under `docs/`.
## Documentation standard
- Describe the current repository, not an abandoned design, planned feature, or remembered implementation.
- Verify file paths, commands, task names, version locations, workflow names, artifact names, and configuration keys against the current tree before documenting them.
- Keep Android and Desktop behavior, versioning, release tags, build commands, and artifact p# Levyra GitHub Automation Instructions
These instructions extend the root `AGENTS.md` for workflows, actions, issue templates, release automation, and repository scripts under `.github/`.
Before editing a workflow, load `.agents/skills/levyra-ci-workflows/SKILL.md` and, when applicable, `.agents/skills/levyra-security-review/SKILL.md` and `.agents/skills/levyra-release-check/SKILL.md`.
## Trust boundaries
- Treat `pull_request`, `pull_request_target`, forks, issue comments, workflow dispatch inputs, do# Levyra Android Module Instructions
These instructions extend the root `AGENTS.md` for every file under `app/`. The root contract remains authoritative; when rules differ, follow the stricter safety or validation requirement.
## Module scope
The Android application is built with Kotlin, Jetpack Compose, AndroidX Media3, Room, WorkManager, OkHttp, Coil, and the existing extractor integration.
Before editing, load every matching native skill under `.agents/skills/` and the referenced canonical Claude/dom{
"mcpServers": {
"jcodemunch": {
"command": "bash",
"args": [
"-lc",
"exec \"$(git rev-parse --show-toplevel)/scripts/claude-jcodemunch-mcp.sh\""
]
}
}
}
---
name: levyra-compose
description: Implement, debug, or review Levyra Jetpack Compose screens, state projections, navigation, animation, lifecycle, accessibility, performance, RTL, localization, edge-to-edge, adaptive layout, and screenshot-validation changes.
---
# Levyra Compose workflow
## Required context
1. Read the root `AGENTS.md` and `app/AGENTS.md`.
2. Read `docs/ARCHITECTURE.md`.
3. Read `.agents/claude/rules/compose-ui.md`, `.agents/claude/rules/localization.md`, and `.agents/claude/rules/a---
name: levyra-security-review
description: Perform an evidence-based Levyra security review and Codex Security workflow covering threat modeling, attack paths, validation, remediation, revalidation, secrets, provider URLs, redirects, SSRF, MIME confusion, permissions, privacy, logging, workflows, dependency changes, update verification, and untrusted input. Use automatically for vulnerability scans, security findings, dependency risk, authentication, trust-boundary changes, sensitive data, or security-re---
name: levyra-project-manager
description: Turn Levyra requirements and roadmap outcomes into one reviewable active phase using docs/project/SPEC.md, docs/project/ROADMAP.md, docs/project/TASKS.md, acceptance criteria, validation, owner checkpoints, and domain-skill routing.
---
# Levyra project management workflow
## Use this skill when
- defining or updating a tracked engineering phase;
- reconciling requirements, roadmap, tasks, architecture, and code;
- preparing a precise implementation plan for ---
name: levyra-pr-review
description: Review a Levyra branch, commit, patch, or pull request for correctness, regressions, concurrency, lifecycle, security, data safety, UI behavior, CI, release risk, missing tests, and merge-readiness evidence.
---
# Levyra pull request review workflow
## Required context
1. Read the root `AGENTS.md` and every nearest `AGENTS.md` covering changed files.
2. Read all canonical `.agents/skills/` and `.agents/claude/rules/` owners relevant to the diff.
3. Inspect the comp---
name: levyra-engineering
description: Coordinate cross-domain Levyra work that spans multiple subsystems, or orient an agent when no single specialized Levyra skill is sufficient. Prefer a specialized levyra-* skill for focused tasks.
---
# Levyra engineering coordinator
Use this skill for repository orientation, architectural investigations, or changes that genuinely cross several Levyra domains. Do not use it as a substitute for a more precise specialized skill.
## Start here
1. Read the root `AGE---
name: levyra-release-check
description: Automatically use for Levyra pre-merge/pre-release validation, Android emulator or physical-device runtime verification, versions, signing, builds, packaging, artifacts, checksums, workflows, secrets, release-note authoring, and truthful manual evidence.
---
# Levyra release validation workflow
## Required context
1. Read the root `AGENTS.md`, `app/AGENTS.md`, `desktop/AGENTS.md`, and `.github/AGENTS.md` as applicable.
2. Read `.agents/claude/rules/testing-rele---
name: levyra-r8-proguard
description: Automatically use for Levyra R8, Proguard, minification, resource shrinking, keep rules, consumer rules, release-only crashes, reflection/serialization/JNI shrinking issues, APK size, mapping files, missing classes, or shrinker configuration review.
---
# Levyra R8 / Proguard workflow
## Why this is a first-class skill
Levyra release builds are minified and resource-shrunk. Shrinker changes can therefore improve size and optimization, but can also create failures---
name: levyra-motion-artwork
description: Implement, debug, or review Levyra decorative motion artwork, provider matching, muted media playback, prefetch, caching, lifecycle, and remote-media security.
---
# Levyra motion artwork workflow
## Required context
1. Read the root `AGENTS.md` and `app/AGENTS.md`.
2. Read `.agents/claude/rules/player.md`, `.agents/claude/rules/security.md`, and the relevant architecture sections.
3. Inspect the motion engine, provider matching, ViewModel ownership, artwork l---
name: levyra-codex-bootstrap
description: Use when Codex tooling, jCodeMunch, RTK, Matt Pocock skills, automatic setup, context efficiency, or Codex MCP/bootstrap behavior for Levyra is being inspected, repaired, or changed. Do not load for ordinary coding tasks unless the bootstrap itself is relevant.
---
# Levyra Codex bootstrap
The repository owner explicitly authorizes Levyra's checked-in Codex bootstrap to install and run only the pinned/allowlisted tooling defined by the repository:
- the exist---
name: levyra-player
description: Implement, debug, or review Levyra playback, queue, Media3, MediaSession, notification, Android Auto, prefetch, audio mode, video mode, and player lifecycle changes.
---
# Levyra player workflow
## Required context
1. Read the root `AGENTS.md` and the nearest applicable `AGENTS.md`.
2. Read `docs/ARCHITECTURE.md`.
3. Read `.agents/claude/rules/player.md` and `.agents/claude/rules/architecture.md`.
4. Inspect the complete current path through user intent, ViewModel/con---
name: levyra-design-taste
description: Automatically use together with the matching Levyra UI skill for any visual redesign, UI polish, visual hierarchy, spacing, typography, color, shape, motion, screenshot/reference recreation, or request to make Levyra feel more premium, modern, distinctive, cohesive, or less AI-generated. Applies to Android Compose and Desktop Compose Multiplatform product UI.
---
# Levyra design taste
## Purpose
This skill is Levyra's product-design quality gate. It exists to pr---
name: levyra-database
description: Implement, debug, or review Levyra Room entities, DAOs, migrations, schemas, caches, stores, backups, downloads, favorites, playlists, history, and persistent queue data.
---
# Levyra database workflow
## Required context
1. Read the root `AGENTS.md` and `app/AGENTS.md` when working in the Android module.
2. Read `docs/ARCHITECTURE.md`.
3. Read `.agents/claude/rules/data-room.md` and `.agents/claude/rules/architecture.md`.
4. Inspect every affected entity, DAO, repo---
name: levyra-android-reverse-engineering
description: Mandatory Levyra workflow for lawful Android artifact decompilation, APK/XAPK/AAB/DEX/JAR/AAR analysis, API extraction, compiled call-flow tracing, obfuscation analysis, and authorized dynamic diagnostics.
---
# Levyra Android Reverse Engineering
Use this skill whenever the task analyzes compiled Android artifacts rather than
Levyra source directly: APK, XAPK, AAB, DEX, JAR, AAR, jadx/smali output,
compiled API extraction, binary call-flow tracing,---
name: levyra-openclaw-orchestrator
description: Coordinate Levyra work through OpenClaw using a dedicated repository workspace, compact delegation, specialized review and CI agents, project-native skills, evidence collection, durable memory, and owner-controlled publication.
---
# Levyra OpenClaw orchestration workflow
## Role
OpenClaw coordinates bounded work. It does not replace Levyra requirements,
domain skills, coding runtimes, CI, independent review, or owner approval.
Use the existing Levyra ---
name: levyra-extractor
description: Implement, debug, or review Levyra InnerTube, YouTube extraction, stream resolution, player-config synchronization, token, retry, timeout, cache, and network fallback behavior.
---
# Levyra extractor workflow
## Required context
1. Read the root `AGENTS.md` and the nearest applicable `AGENTS.md`.
2. Read `docs/ARCHITECTURE.md`.
3. Read `.agents/claude/rules/extractor-network.md` and `.agents/claude/rules/security.md`.
4. Inspect the resolver, clients, runtime confi---
name: levyra-desktop
description: Implement, debug, or review Levyra Desktop for Windows, including Kotlin/JVM, Compose Multiplatform, libvlc, queue, downloads, persistence, mini player, deep links, updates, packaging, and desktop releases.
---
# Levyra Desktop workflow
## Required context
1. Read the root `AGENTS.md` and `desktop/AGENTS.md`.
2. Read `desktop/README.md` and the relevant architecture documentation.
3. Inspect the affected `desktop/core`, `desktop/player`, and `desktop/app` code plus n---
name: levyra-android-intent-security
description: Automatically use for Levyra Android Intent, deep-link, PendingIntent, exported component, receiver, service, provider, URI-grant, FileProvider, caller-verification, or onNewIntent security work. Pair it with levyra-security-review and the affected Android domain skill.
---
# Levyra Android Intent security workflow
## Purpose
This is Levyra's Android-specific component-boundary security layer. It adapts
selected practices from Google's `android/skills---
name: levyra-ci-workflows
description: Automatically use for Levyra GitHub Actions, CI, F-Droid, Gradle/AGP/Kotlin/KSP compatibility, build performance, configuration/build cache, artifacts, release automation, workflow security, or configuration-sync work.
---
# Levyra CI and workflow workflow
## Required context
1. Read the root `AGENTS.md` and `.github/AGENTS.md`.
2. Read `.agents/claude/rules/testing-release.md` and `.agents/claude/rules/security.md`.
3. Inspect every workflow, reusable action, s---
name: levyra-android-performance
description: Automatically use for Android runtime performance investigations involving Perfetto/System Trace, jank, latency, startup, CPU scheduling, blocking, memory, I/O, IPC, graphics, power, or measured frame/runtime bottlenecks. Pair it with the affected Levyra domain skill.
---
# Levyra Android performance workflow
## Purpose
This skill is the Android-specific evidence layer for runtime performance. It
adapts the strongest parts of Google's official `android/sk---
name: levyra-real-engineering
description: Route non-trivial Levyra engineering through the Matt Pocock real-engineering workflow. Use automatically for features, architectural changes, bugs, test/build failures, regressions, unexpected behavior, multi-step work, or requests that risk AI slop when root-cause investigation, specification, focused implementation, or independent review is useful; skip the ceremony for tiny obvious edits.
---
# Levyra real-engineering workflow
This skill adapts Matt Pococ---
name: levyra-context-efficiency
description: Use for genuinely high-volume Levyra work such as builds, tests, lint, logs, broad searches, dependency output, Git/GitHub or CodeRabbit inspection, CI diagnostics, agent setup, cross-domain exploration, or useful cross-session retrieval. Reduce token waste without reducing engineering rigor.
---
# Levyra context-efficiency workflow
## Purpose
Spend model context on code and decisive evidence. This skill reduces repeated
instructions, broad reads, noisy co---
name: levyra-mode
description: Use only for explicit owner execution cues such as VAI, PROCEDI, INTERVIENI, FAI TU, FALLO TU, or requests to open/create a PR. Keep execution fast and low-noise while preserving full Levyra engineering rigor.
---
# Levyra mode
## Purpose
Make owner-directed work faster to execute and easier to follow. Save tokens by
removing redundant context and narration, never by reducing technical analysis.
Root/scoped `AGENTS.md`, always-on guards, domain skills, validation, revie---
name: levyra-humanizer
description: Apply a compact final prose pass to Levyra pull request descriptions, release notes, and owner-facing writing so it sounds natural without changing facts, evidence, structure, or validation state.
license: MIT
metadata:
version: "2.11.2-levyra-compact"
upstream: "blader/humanizer@e2e92e7b4b8229253ed5c8e81dc65463fdeddda5"
---
# Levyra humanizer
## Goal
Make finished prose sound like a person wrote it while preserving every factual
claim. This is a lightweight finone