Skip to content
Agentmart
Submit a tool

android

Android project audit and preflight skill.

Skill by FelixGraeber

IndexedMIT★ 3
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/felixgraeber-android

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Android project audit and preflight skill. Builds a shared audit_context.json from static project evidence, then uses specialist agents to interpret that evidence with explicit confidence and deterministic gate caps. Supports Kotlin/Compose and XML/Java repos, including library-only projects. Triggers on: "android audit", "android assessment", "android review", "android project health".

Use cases: security

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found
Runs as
Python package

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/android

What's inside

Skill name
android
Description
Android project audit and preflight skill. Builds a shared audit_context.json from static project evidence, then uses specialist agents to interpret that evidence with explicit confidence and deterministic gate caps. Supports Kotlin/Compose and XML/Java repos, including library-only projects. Triggers on: "android audit", "android assessment", "android review", "android project health".
License
MIT
Instructions
188 lines in SKILL.md · view on GitHub

Before installing

  • Ships executable scripts

Files (24)

  • SKILL.md
  • android-16-changes.md · references/
  • compose-best-practices.md · references/
  • material-design-3.md · references/
  • owasp-mobile-2024.md · references/
  • play-store-policies.md · references/
  • quality-gates.md · references/
  • scoring-weights.md · references/
  • vitals-thresholds.md · references/
  • requirements.txt
  • facts.json · rules/
  • rules.json · rules/
  • audit_context.schema.json · schemas/
  • finding.schema.json · schemas/
  • analyze_compose.py · scripts/
  • analyze_dependencies.py · scripts/
  • analyze_gradle.py · scripts/
  • analyze_manifest.py · scripts/
  • build_audit_context.py · scripts/
  • check_r8_config.py · scripts/
  • common.py · scripts/
  • render_rule_docs.py · scripts/
  • scan_project.py · scripts/
  • score.py · scripts/

SKILL.md (first lines)

# Android Project Audit Skill

## Quick Reference

| Command | Description |
|---------|-------------|
| `/android audit [path]` | Full multi-category audit using a shared evidence bundle |
| `/android architecture [path]` | Architecture and module structure review |
| `/android performance [path]` | Static performance preflight |
| `/android security [path]` | OWASP-oriented security audit |
| `/android compat [path]` | Android 15/16 compatibility preflight |
| `/android design [path]` | Design system implementation audit |
| `/android accessibility [path]` | Accessibility preflight from static evidence |
| `/android testing [path]` | Testing strategy and risk coverage review |
| `/android build [path]` | Build system and dependency hygiene review |
| `/android playstore [path]` | Play preflight from source evidence |
| `/android plan [path]` | Prioritized improvement roadmap |

## Consolidated Skill Routing

Use six public skills. Keep the detailed former specialist guides as internal
references so no review knowledge is lost:

- `android`: orchestration, full audit, and planning. Read
  `../android-audit/GUIDE.md` or `../android-plan/GUIDE.md` when needed.
- `android-engineering`: architecture and build system.
- `android-quality`: performance and testing.
- `android-ui`: design and accessibility.
- `android-release`: platform compatibility and Play readiness.
- `android-security`: security.

Do not invoke a `GUIDE.md` file as a separate skill.

## Evidence Contract

The canonical static evidence contract is:

- `schemas/audit_context.schema.json`
- `schemas/finding.schema.json`

The canonical rule registry is:

- `rules/rules.json`
- `rules/facts.json`

Generated markdown references:

- `references/quality-gates.md`
- `references/scoring-weights.md`

Agents should consume `audit-context.json` instead of re-parsing the repo freehand.

## Project Detection

Confirm an Android project by scanning for at least one of:

1. `**/build.gradle.kts` or `**/build.gradle`
2. `**/settings.gradle.kts` or `**/settings.gradle`
3. `**/src/main/AndroidManifest.xml`

If none are present at `[path]`, stop and report that no Android project was detected.

## Project Classification

Classification is deterministic and happens before agent dispatch.

Ordered checks:

1. `sdk-library` if there is no application module and at least one Android library module
2. `single-module` vs `multi-module` from discovered modules
3. `compose-first` vs `xml-legacy` vs `hybrid` from `src/main` sources only

The resulting `project_type` is stored in `audit-context.json` with:

- `repo_kind`
- `app_shape`
- `ui_stack`

## Orchestration Flow

### Step 1: Scan Structure

Run:

```bash
python skills/android/scripts/scan_project.py [path] --json
```

### Step 2: Extract Static Evidence

Run:

```bash
python skills/android/scripts/analyze_gradle.py [path] --json
python skills/android/scripts/analyze_manifest.py [path] --json
python skills/android/scripts/analyze_compose.py [path] --json
python skills/android/scripts/analyze_dependencies.py [path] --json
python skills/android/scripts/check_r8_config.py [path] --json
```

### Step 3: Build Shared Audit Context

Run:

```bash
python skills/android/scripts/build_audit_context.py [path] --output generated/audit-context.json
```

Agents receive `generated/audit-context.json` as their primary input.

### Step 4: Parallel Agent Interpretation

Dispatch relevant agents with:

- project root
- `generated/audit-context.json`
- selected category

Agents must:

From FelixGraeber/claude-android-audit-skill, MIT.

Signals

Stars
3
Forks
0
Open issues
0
Last commit (this folder)
Jul 29, 2026 (3 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from FelixGraeber

Cite this page

android skill. Agent Mart. https://agentmart.dev/skills/felixgraeber-android (last updated 2026-10-24)