Local agent infrastructure in one stdlib-only Go binary

A zero dependency lightweight static analyzer + CVE scanner designed for adversarial-shape code in python to detect supply chain attacks before they reach your interpreter.

Zero-dependency, single-file Python implementations of popular libraries — benchmarked for performance parity | 零依赖单文件 Python 常用库实现，性能对标主流库

HUMMBL MCP Server

Plumego is a standard-library-first Go HTTP toolkit for explicit, maintainable, agent-friendly services, with zero external dependencies. It is designed to be simple, elegant, and efficient, making it ideal for small to large projects.

LLM-neutral skill that fetches a URL to a temp file so agents can pipe the path through rg / jq / awk instead of dumping the whole page into context. Same shape as pi-webfetch / opencode webfetch. Single-file Python 3 stdlib. Works on macOS, Linux, Windows.

TypeScript agent orchestration runtime for HUMMBL — kernel, router, adapters, control-plane, governance gateway. Multi-agent coordination with Base120 mental models.

Plan Mode orchestration tool — wraps Claude Code with structured Plan → Decompose → Execute workflow

Cross-platform CLI for the Linode API plus a Claude skill that drives it under explicit AI-safety constraints. Token never enters AI context.

Headless Chromium on ARM without Playwright, X11, or API keys. Termux, Raspberry Pi, embedded Linux. Public domain.

Runtime safety primitives for AI agent systems — kill switch, circuit breaker, delegation tokens, governance bus. Zero dependencies. 58 tests. 7/10 OWASP ASI coverage.

A plugin for Hermes Agent for meal planning and refrigerator management. It suggests recipes based on your ingredients and generates optimized shopping lists, combining conversational AI with a deterministic core without dependencies.

HUMMBL Bibliography — provenance corpus, BibTeX citations, and position papers for Base120, HCC, BKI, and AI governance research.

Pre-dispatch underwriting witness for delegated agent task slices: is this slice covered by an attested trust envelope and sufficient liability headroom? stdlib-only, deterministic, hash-chained decision ledger.

Check what scopes your GitHub PAT actually has + what API calls it can/can't make. Stop guessing why 200 OK quietly does nothing.

AI deployment policy drift tracker

Edge robotics control evaluation engine checking telemetry deviations and sandboxed detour plans

Operational monitor for multi-agent AI. Per-action observations, per-agent rolling trends, baseline-relative alarms, named recommendations.

Single-file Python disk cache for LLM API calls. Wrap any provider client. Stop paying for the same call twice during dev/eval/demo.

Persistent LRU cache with atomic writes, portable file locking, zero dependencies. Drop-in replacement for functools.lru_cache that survives restarts.