Runtime firewall for MCP and tool-augmented agents — descriptor integrity, static vetting, per-call policy, taint tracking, consent ledger

The transport-neutral agent tool contract for the Voyant framework (voyant#2792).

Normalize URLs with PostCSS

ggui protocol types — events, renders, WebSocket, MCP, LLM models

Squad SDK — Programmable multi-agent runtime for GitHub Copilot

Production-ready TypeScript AI Agent Runtime

Trace imports, detect unused code, clean dependencies — all with a super-fast CLI

JavaScript build tool, similar to Make or Rake

Policy enforcement point service for applying decisions to protected data flows.

Eslint plugin checking architecture boundaries between elements

AI Agent Governance for TypeScript — policy enforcement, scoring, compliance, and audit for AI agents

Fail-closed per-turn step, tool-call, and provider-token budgets for DeepSeek Harness

OpenCode plugin for token-budget guardrails: counts budgeted tool calls, injects reminders at thresholds, audits usage automatically, and makes session handoff a single tool call

A small collection of ESLint rules to improve usage of id attributes in your markup.

Provider connections for the Axon agent runtime — Anthropic, OpenAI, Google, Groq, Ollama and more.

Know what every AI call really costs — a live ledger of your agent's real spend, plus a hard-stop before runaway spend crosses your ceiling. Vercel AI SDK middleware + LiveKit / Vapi / Retell voice adapters. Local, no account, no telemetry.

MCP server that turns your design guidance into a checkable process -- evaluates UI components from external libraries (shadcn/ui, 21st.dev, ReUI) or your own registered design system against a requirements checklist, then tells the agent whether to reuse

An agent-friendly LLM proxy that cleans up Kimi K2 output, fixes tool calls, and normalizes responses for OpenAI/Anthropic-style clients.

CLI tool for Angular

Opt-in quota tracking and enforcement for Loopstack workflows