Agent Skills by pepsi1978
Showing 21 vetted skills indexed across 1 GitHub repositories.
cross-platform
Align APIs, paths, and behavior across macOS Swift and Windows C# apps.
android-audio
Implement Android audio with SoundPool, Media3, and Oboe APIs.
auto-verify-iterate
Automate end-to-end verification and iterative improvement after coding tasks.
android-development
Scaffolds Kotlin Android projects with Jetpack Compose, MVVM, and multi-module structure.
hook-forge
Creates cross-platform Claude Code hooks with automated validation and logging for PS1 and SH environments.
codebase-memory-reference
Reference Codebase Memory MCP tools, edge types, and graph query patterns.
undo-changes
Revert local and GitHub commits using git reset or revert.
tampermonkey-version
Bump @name and @version headers in Tampermonkey .user.js files before commits.
sound-search
Search and retrieve audio clips from Freesound and Zapsplat for app use.
codebase-memory-tracing
Traces function call paths across modules and services to reveal callers and dependencies.
codebase-memory-exploring
Index a repository into a graph and query code elements.
codebase-memory-quality
Identify dead code and refactor candidates using static graph analysis.
direktiven-recherche
Coordinates five parallel web researchers to find new or improvable directive proposals.
string-extraktor
Extract hardcoded German strings from Android Kotlin/Compose and XML layouts into string resources.
rechtssicherheit
Audit Android app legal compliance for Google Play releases.
übersetzung
Translate Android strings.xml into 26 locales with per-language validation.
agent-briefing
Generate a 10-line situational-awareness briefing with branch, commits, issues, and file ownership.
superintelligenz
Guide autonomous self-improvement and architectural decisions using an eight-dimension intelligence framework.
app-monetizer
Audits Android apps' monetization potential and generates a Markdown report with ready-to-run prompts.
resilient-bugfixing
Automate bugfixing with root-cause analysis, defensive patching, and validation.
selbstbeobachtung
Automate post-task retrospectives and persist learnings to MEMORY.md and rules.