Signatrust SDK — a verifiable, tamper-evident receipt for every AI decision. Includes local hashing, Decision Boundary Disclosure (DBD), Deployment Evidence Profile pinning, human-review attestation, and a customer-side ReceiptReferenceStore.

DSH Web taskboard plugin with workspace-scoped Agent sessions, staged human review, delivery confirmation, and durable archival

Model Context Protocol server for Nodl — give your AI client a live map of your system: areas, sub-areas, integrations, databases, and the change graph. Document features back into the Nodl inbox for human review.

Screen a person or company name against Nigeria's official NIGSAC sanctions register as an MCP tool — a best-effort, diacritic-folded, transliteration-tolerant screening signal for human review. Live register, keyless.

Claude Code PreToolUse hook that hard-gates dangerous shell commands before they ever run, with a double-layer degrade counter (same-issue repeated -> 'ask' human review; cumulative -> hard 'deny'). Zero DSH/runtime deps; cross-platform. Ports the deny +

Transform unstructured data to structured data using LLMs with human review

Pluggable executor workflow engine for AI coding agents. Dispatch workflow steps to Claude Code, Codex CLI, or human review checkpoints.

TLDR, a local-first human review and reading workspace for structured artifacts. Provides the tldr binary.

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

ConfidenceIndicator — a model-confidence meter with a human-review threshold — Nucleux agentic UI for React.

Local-first long-running bug fixer: staged AI patches behind hash-bound human review.

ReachOps - outreach research tool. Builds a database of US companies with hiring signal (Form 5500 backbone), resolves domains, discovers job boards and postings, surfaces hiring stakeholders, and drafts outreach emails for human review and send.

Official ClearedBy SDK — gate agent actions, wait for human review, guard a function.

Amazon A2I integrates human review for low-confidence ML predictions.

Your first mate: an unattended daemon that drains ready-for-agent Linear tickets, implementing each in an isolated git worktree via claude -p and opening a PR for human review.

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

Task-by-task execution with human review checkpoints for early error detection

Human-in-the-loop approval gates for AI agents — pause function calls for human review, then resume or abort.

Deterministic finance operations AI workflow example with human review, fixed rules, and synthetic data only.

Lets you prompt users to submit app store ratings and reviews without leaving your app - using platform native dialogs.