Browse agent skills

Automated prompt injection & jailbreak testing against LLM system prompts using local edge models for CI/CD pipelines

A private, fully local, LLM-powered file renaming solution.

A Swarmauri tool for ranked image-vector retrieval from images or embeddings.

Swarmauri chat and tool-calling adapters for NVIDIA NIM OpenAI-compatible deployments.

OpenRouter models for Swarmauri chat, tool-calling, vision-language, and image-generation workflows.

Provider-native Swarmauri chat and tool-calling adapters for the xAI Inference API.

One character image → a full pack of animated chat stickers (WeChat / Telegram / LINE / Discord). Claude Code & Codex skill · 给 AI 一张图，做一整套会动的表情包

My personal C# styling and .NET preference with my frequently used library and Sakura Framework usage guide

Provider-native Swarmauri chat and tool-calling adapters for Azure OpenAI.

OpenAI Realtime and persistent Responses WebSocket model adapters for Swarmauri.

Aggregated OpenAI provider resources for Swarmauri models.

Intrinsic Attention Bottlenecks for Vision Transformers & LLM Prompt Compression.

A measure-first layer that reduces token use of LLM requests: token accounting, exact repeated-context removal, bounded tool-output trimming, an exact-match cache that needs a tenant scope, retries and fallback, and an evaluation harness that compares quality before and after. Standard library plus ragleap-security. Every step reports what it changed and whether it can lose information. Not independently audited.

No description available

High-fidelity multi-format document parsing and chunking engine for RAG pipelines

Skills that stop AI agents from writing backend code that passes the demo and breaks in production. Critical mistakes: 25/150 runs without, 2/150 with. Claude Code, Codex, Cursor.

Evidence-based mobile app auditor for Claude

Claude Code skill: turn raw talking-head footage into a polished 9:16 reel, with a pop-out face panel, captions, B-roll and sound effects.

dia (嗲) — lightweight agent framework: knowledge index, audit, session ingest

Scores Claude Skills 0–100 with plain-English fixes, catches trigger collisions, and automatically re-validates skills when they change.

O Claude Code com método: pensa antes de construir e comprova antes de dizer que terminou. Em português, com 3 níveis de cuidado. Baseado no Superpowers.

Propostas técnicas e comerciais com a sua marca, feitas pela sua IA: escopo preso ao pedido, âncora de valor, PDF e link com código de acesso na Cloudflare.

Configurable sync and async Python clients for LLM and media APIs

Local proxy & TUI that catches 1-token prompt/KV cache prefix misses and auto-heals requests for Anthropic, OpenAI, llama.cpp and Ollama.

Extreme-ownership agent teams for Claude Code: bounded subagents with hook-enforced authority, MissionBrief → UpwardReport contracts, and an outcome ledger.

Bansat (بنسټ) — lay the foundation before you build. A blueprint-first framework and Claude Agent Skill by Muhammad Tufail Khan

A Claude skill that turns your course slides into an exam prep system: a study plan, a mastery-gated quiz app, and an optional cheat sheet.

Free agent skills for TransformPipe: publish Markdown as a page, Confluence and Notion exports to Markdown

Drop-in audit logging for LLM API calls (OpenAI, Anthropic)

Search and verify businesses on FirmLedger from n8n workflows.