Intelligence, earned.

Agent harness skills/instincts directly applicable to BLC Claude Code workflow; MIT license. Triage verdict: ADOPT.

Agentic skills framework with development methodology — Claude Code skills. Triage verdict: ADOPT.

Long-horizon SuperAgent harness with memory/skills/subagents — orchestration patterns for Jacer. Triage verdict: ADOPT.

Research skill covering Reddit/X/YouTube/HN/Polymarket — directly useful for MetaEngine Michael. Triage verdict: ADOPT.

Official Anthropic Python SDK. Used for cloud tier (complex reasoning, code-gen) in JARVIS model routing.

Fast async PostgreSQL client. Used as the DB driver for JARVIS RIDB FastAPI service.

Multi-agent conversation framework by Microsoft. Provides GroupChat, AssistantAgent, UserProxyAgent abstractions. Key pattern source for JARVIS multi-agent orchestration.

Role-based multi-agent framework with crew/task/agent primitives. Useful for structured task decomposition — maps directly to JARVIS execution pipeline steps.

Modern Python web framework used as JARVIS RIDB executor API foundation. Async, pydantic-native, OpenAPI auto-docs.

Autonomous AI research agent that performs deep web research and generates structured reports. Uses planning + scraping + synthesis pipeline, suitable as a research backbone for any JARVIS task requiring information gathering.

Async HTTP client. Used for Michael dispatcher and Hades Ollama calls in JARVIS.

Universal LLM API router — single interface for OpenAI, Anthropic, Ollama, and 100+ providers. Key for JARVIS model routing layer (local/cloud 70/30 split).

Structured logging library with file rotation and async support. Used for JARVIS execution logs per run.

Article scraping and NLP library. Extracts full text, authors, publish dates from news URLs. Michael can use this for Nepal Elections content ingestion.

Local LLM server running on JARVIS (nomic-embed-text) and Hades. Already deployed — adapt for embedding tier and local reasoning.

Python client for pgvector. Used for embedding insert/query in RIDB.

Data validation library. Used throughout JARVIS for structured output contracts and config validation.

Environment variable loader from .env files. Used to load ANTHROPIC_API_KEY and other secrets in JARVIS.

Beautiful terminal output library. Used for JARVIS CLI run reports and structured logging.

Python-based SEO automation tool for article optimization, meta generation, and keyword density analysis. Wraps common SEO checks into a CLI pipeline. Integrates well with content agents like Michael for Nepal Elections and similar editorial workflows.

SQLAlchemy + Pydantic hybrid ORM. Used in JARVIS RIDB for type-safe DB access.

Web scraping focused on text extraction and metadata. Faster than newspaper4k for bulk pipelines. Good for content-at-scale use in MetaEngine.

CLI framework built on Click. Used for JARVIS management CLI (seed, ingest, run).

Multi-agent LLM trading framework — highest priority for botmeg-advisor patterns. REFERENCE only during live window. Triage verdict: ADOPT (blocked).

Distributed task queue. Reference for async job patterns — RIDB may need background embedding backfill queue.

Lightweight vector database. REFERENCE only — JARVIS already uses pgvector, but chroma patterns useful for client-side embedding work on Jacer.

Production crypto trading bot with strategy framework and backtesting. Reference only — no engine writes during BotMeg live window.

Foundational LLM framework with chains, tools, memory, agents. Reference for prompt engineering patterns and tool-use abstractions. Already widely adopted.

RAG framework with document loaders, vector store integrations, query engines. Reference for retrieval patterns when querying RIDB.

Open-source software engineering agent that can browse the web, write code, and execute commands in a sandboxed environment. Reference for agentic code execution patterns.

Multi-agent trading system with analyst/researcher/trader roles. Reference for BotMeg architecture. DO NOT adapt without live-window clearance.

High-throughput LLM inference server. Reference for future JARVIS GPU inference if VPS gets upgraded with GPU.