Framework-independent, fail-closed human approval middleware for consequential AI agent tool calls.

ConnectWise MCP server for LLM agents — Claude, local Qwen via LM Studio, Cursor, Codex. PSA Lite free read-only desk + full PSA/RMM tool calling with human approval. Self-hosted, BYOM. Private beta (public soon): pulsatrix.ca

Deny-by-default policy gate for AI agent tool calls: allowlists, argument validation, rate caps, human-approval hooks, dry-run mode, and an audit trail. Framework-agnostic, zero dependencies.

Manages requests that require user approval

Approval Tests Library - Capturing Human Intelligence

Record a captioned demo video of any web app with one command — then grow into autonomous CWS/X/YouTube launch assets with final-file QA and a human approval gate.

Human-readable error messages for Ajv (Another JSON Schema Validator).

Reins (驭程) console — the Vue 3 web frontend of the Reins engineering platform: live flow DAG, traceability matrix, human approval gates, and agent activity streams over DeepSeek Harness sessions. This 0.0.1 release is a name reservation; the first functio

Declare your domain ontology once — get agent-facing docs and runtime enforcement (typed MCP server, human approval, audit log). Under active development.

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

User-facing permission presets (ctx.permissionPresets) for the DeepSeek Harness: one product-level Permissions select bundling the sandbox-mode and approval-policy knobs, written through to their own session events

pendpost is a free, open-source (MIT), local-first social media planner where an AI agent drafts and schedules posts across Instagram, Facebook, LinkedIn, YouTube, X, Telegram, Discord, Mastodon, and Nostr, plus long-form blogs on WordPress and Ghost, wit

Interven AI firewall — TypeScript SDK. Scan agent tool calls before they execute. Block malicious requests, redact PII/secrets, route risky actions to human approval.

Run the CodyRelay relay on your own machine: multi-agent coding loop with human approval gates. Your code and AI subscription never leave your computer.

Dench agent workspace CLI.

Generate human-readable ids from lists of easy-to-spell nouns and adjectives

Converts a number to/from a human readable string: `1337` ↔ `1.34kB`

Interven JS/TS SDK — Approvals + Audit for AI Agents. Scan tool calls through policy + risk scoring. Block, redact PII, or route to human approval via Slack.

Component testing utils for Vue 3.

Pi extension for resource-aware security policy engine and tool authorization