L

llm-agents

Zambo: the cross-AI execution layer. 100+ native tools over one MCP endpoint, with a verifiable execution receipt for every call (AER-1 open draft). Free: 20 calls/tool/day, no account. Day Pass $0.99 USDC on Base for 24h; Zambo Pass $29/month ($19/month for verified $ZAMBO holders). https://zambo.dev

An explainable, offline-first, agent-native URL inspector — detects deceptive URLs and tells you exactly why, with no network

Open-source computational laboratory for commons dynamics, governance, Seneca cliffs, and generative agents. Five-paper series (BDPD⁰ ...) + mini-course + research notes + annotated bibliography on Zenodo. Node.js simulation platform + Python LLM agents + The Forest of Humbaba card game.

Poetra AI – AI-Powered WhatsApp Online Store

Developed an AI-powered WhatsApp chatbot that enables customers to interact with an online store through natural language. The system helps customers discover products, manage their shopping cart, and proceed through the checkout process directly within WhatsApp.

Key contributions and features:

Developed an LLM-powered conversational interface for product discovery and customer assistance. Implemented product search, category browsing, cart management, and checkout workflows. Integrated GLM-4.5-Air through OpenRouter for natural-language interactions. Implemented Retrieval-Augmented Generation (RAG) to provide the AI with relevant product information. Integrated Midtrans for payment and checkout processing. Connected the system with WhatsApp for real-time customer interactions. Built an administrative dashboard using Vue.js and MySQL for managing products, customers, orders, users, and conversation logs. Implemented API integrations between the AI layer, database, WhatsApp service, and payment system. Conducted Black Box and White Box testing to validate system functionality and reliability. Designed the system to support real-world e-commerce workflows from product discovery through payment.

Kevlar Benchmark: OWASP Top 10 for Agentic Apps (AI-Agents) 2026 a Red Team Benchmark.

Self-hosted video API in one Docker container. Lipsync (LatentSync 1.5, Wav2Lip), face restoration (GFPGAN), ffmpeg ops (trim/concat/transcode/scale/mux/extract/thumbnail/gif), ffprobe metadata, async jobs + webhooks. REST + MCP. CPU + CUDA. Drive from shell, generated Go/Python clients, or LLM agent.

Self-hosted audio API in one Docker container. Stem separation, mastering, MIR (BPM/key/chords), audio→MIDI, text-to-music + SFX (MusicGen/Stable Audio/Riffusion/AudioLDM2), diarization, speech enhancement, restoration, embeddings + zero-shot tagging, workflows, async + webhooks. REST + MCP. CPU + CUDA.

A series of incremental example usages of promptfoo and its application into LLM related features

A self-improvement flywheel for Claude Code — 4 layers that turn ephemeral sessions into compounding, durable learning.