Seald reversibility tools

MCP server — reversibility intelligence for autonomous AI agents

Context compression plugin for opencode — port of headroom with Kneedle adaptive-K, BM25 relevance scoring, and CCR reversibility

Sagaz — undo for AI agents: an MCP proxy with a hash-chained effect ledger, reversibility classification, policy gates and preview. Real release coming soon.

Context compression plugin for opencode — port of headroom with Kneedle adaptive-K, BM25 relevance scoring, and CCR reversibility

Phosra MCP server — expose 79 child-safety tools (incl. federation, streaming, reversibility) to any MCP-compatible AI client

EyeOn Center Map v2 — core, viewer, and editor

The AbsoluteJS package manifest contract. Every @absolutejs/* package exports a typed manifest (settings schema, env requirements, adapter slots, wiring recipes, AI tools) from its ./manifest subpath; this package is the contract those manifests are writt

CLI tool for the Lore protocol — structured decision context in git commits

Command line arguments parser and stringifier

Changeset library incorporating an operational transformation (OT) algorithm - for node and the browser, with shareJS support

The crew that builds while you're gone: walk-away, CI-dispatched engineering programs with a decision ledger instead of gates

One unified API for 18 live-chat & customer-support widgets — Intercom, Crisp, Tawk.to, Zendesk, HubSpot, Chatwoot, LiveChat, Drift, Freshchat, Olark, Userlike, HelpScout, Smartsupp, LiveAgent, Gist, JivoChat, Tidio, Sendbird. Zero deps, tree-shakeable, S

API-key pre-action risk gate: any irreversible agent action (money, SQL, delete, email).

Web GUI console for MCP servers: manages the official @deepseek-ai/dsh-mcp-client at runtime — add/edit/enable/disable/reconnect/delete servers, live status, mcpServers JSON import. No MCP protocol code: connection and tool registration are 100% the offic

The Sapience Suite transforms OpenClaw from a reactive assistant into a proactive agent with genuine autonomy. It learns when to act, when to propose, when to ask, and when to explore — calibrated to your actual preferences, not a static policy you had to

One command bootstraps an AI coding agent's working environment in any repo.

Universal LLM privacy gateway — quantum-resilient PII masking, FPE format preservation, and real-time streaming de-masking

18 AI personas deliberate your hardest decisions — structured multi-round deliberation with model diversity across providers.

The TypeScript client for the CAI score API — a typed wrapper over the API's own OpenAPI contract, so a misspelled answer value is a compile error rather than a runtime 400.