Merge autonomy for AI-authored code. Turn your repo's standards into evidence-backed readiness reports the agent reads back.

Logged per-agent plan mode with deployment guidance, a direct slash command, and a user-reviewed exit

Action ledger schema, append-only write helpers, and idempotency primitives for Voyant control surfaces.

Code-first durable orchestration for nanobpmn (ADR 0044/0045). Author a durable workflow as a declarative flow — `defineFlow(id, w => { w.run(...); w.signal(...); w.task(...) })` — and the SDK derives the executable BPMN model, the job types, and the mess

Web Session-log export command and shared download dialog

User interface primitives for console applications

Automation-only Agent Client Protocol server for driving DeepSeek Harness agents over JSON-RPC stdio

Codex-style auto-approval for DeepSeek Harness. An independent reviewer decides allow/deny, fail-closed, with a rationale for every decision. Engine: an LLM route, or TypeSafe's Jev (`reviewer.engine`). · DSH 插件：独立复核器裁决，引擎可选 LLM 路由或 Jev。

No description provided.

React email surfaces for AppKit chat and admin workflows

Multi-agent workflows with full observability and time travel: watch every step live, rewind, fork, and replay any run. Claude Code, Codex, Gemini, any model or harness.

Command gate policies on the devflow transition waterfall

RIVO local MCP — holds the agent's own (non-custodial) wallet, signs each payment, and relays it to the hosted RIVO app for policy + human approval + audit before broadcast. RIVO never holds the key or custodies funds. Zero code: add it to your harness an

Multi-agent social content pipeline that does not break in production: research, brand-voice scripts, NL-directed video edits, publish, analytics and redesign in a closed loop, with per-handoff evals and a human approval gate

A controlled workflow for agentic software development.

n8n community nodes for Guardian Safety Gate — HMAC-signed policy enforcement, atomic execution claims, and automatic human approvals for AI workflows.

Otorga - a consent layer for credentials. A local-first, self-hostable secrets broker that gates every secret access behind hardware-backed human approval, with full process-tree visibility and a tamper-evident audit trail. Under construction.

Agent Approve plugin for OpenClaw - approve or deny AI agent tool calls from your iPhone and Apple Watch

AWS SDK for JavaScript Sagemaker A2i Runtime Client for Node.js, Browser and React Native

User-approval seam (ctx.approval) for the DeepSeek Harness: one-shot permission decisions dispatched to composed answerers over the approval/request waterfall, fail-closed by default