Agentic AI Security

Zedly Shield: Runtime Safety for Agentic AI

Evidence and enforcement layer for OpenClaw. PII redaction, policy-based tool blocking, human approval gates, and tamper-evident audit logs.

For engineering leads, security teams, and compliance officers deploying agentic AI on sensitive workflows.

How Shield Hooks Into the Agent Lifecycle

Every tool call passes through the same five-step pipeline.

Agent Requests Tool Call

LLM decides to run exec, read, write, or HTTP

before_tool_call Hook Fires

Shield intercepts before execution

Policy Engine Evaluates

Rules checked against tool + args

Allow, Block, or Redact

Dangerous calls blocked, PII stripped

Log Tamper-Evident Event

SHA-256 hash-chained to previous event

Cloud Dashboard for Your Fleet

One place to see what every Shield instance is doing. View session runs, event timelines, policy blocks, and PII redactions across all deployments. Push policy changes from the dashboard; the plugin pulls them on its next poll.

  • Fleet-wide stats: instances, events, blocks, redactions
  • Session run table with status, duration, and tool breakdown
  • Event timeline filterable by type and session
  • Policy editor with save and sync to instances
  • API key management and instance provisioning

Dashboard screenshot coming soon.
Join the waitlist for early access.

How It Compares

What you get with Shield vs. the alternatives.

Capability No Protection DIY Middleware Zedly Shield
PII redaction (email, SSN, credit card) None Build your own regex Built-in, on by default
Tool-call blocking None Custom hook code Policy-as-JSON rules
Human approval gates None Build approval queue before_tool_call interception
Audit trail integrity Terminal scrollback Append-only log file SHA-256 hash chain
Multi-instance dashboard N/A Build from scratch Fleet view, per-instance drill-down
Install & upgrade N/A Manual deployment openclaw plugins install
Policy-as-code None Hardcoded logic JSON config, cloud-synced

Get Started

Create your Shield instance, install the plugin, and see events in the dashboard within a minute.

1 Create your Shield instance

Sign up for free, open the Shield dashboard, and create a new instance. Copy the API key it generates.

2 Install the plugin

openclaw plugins install zedly-shield

Requires Node.js 22+ and OpenClaw CLI installed.

3 Add your API key

{ "plugins": { "zedly-shield": { "apiUrl": "https://zedly.ai/api/shield", "apiKey": "sk_shield_paste_your_key_here" } } }

Merge this into your openclaw.json. All protections (PII redaction, shell blocking, audit logging) are on by default. Only add extra keys if you want to disable something.

4 Restart the gateway

openclaw gateway

Look for Zedly Shield starting in the output. Events will appear in your dashboard immediately.

Upgrade later

openclaw plugins update zedly-shield

New protections are enabled by default on upgrade. No config edits needed.

Frequently Asked Questions

What is Zedly Shield?
Zedly Shield is an evidence and enforcement layer for agentic AI workflows. It hooks into agent runtimes like OpenClaw to redact PII before model calls, block dangerous tool calls via policy, gate sensitive actions behind human approval, and produce tamper-evident audit logs with SHA-256 hash chain integrity.
How does Shield integrate with OpenClaw?
Shield is an OpenClaw plugin installed via npm. It hooks into the OpenClaw Gateway lifecycle events (before_tool_call, tool_result_before_model, tool_result_persist, agent_start, agent_end) to intercept, evaluate, redact, and log every action. Install with openclaw plugins install zedly-shield and add it to your openclaw.json.
Does Shield work with other agent frameworks?
Shield is built OpenClaw-first, using its plugin system and hook model. The plugin architecture is designed to support additional runtimes in the future, but OpenClaw is the only supported runtime today.
What data does Shield send to the cloud?
Shield sends event metadata (event type, tool name, session ID, timestamps, block/redact counts) to the Zedly cloud dashboard. It does not send prompt content, tool output, or document data. PII redaction happens locally before any data leaves the host. The cloud dashboard is optional; all events are also logged locally in JSONL format.
Is there a free tier?
The Shield plugin itself is open source and free to use. Local policy enforcement, PII redaction, and JSONL audit logging work without a cloud account. The cloud dashboard (fleet management, cross-instance visibility, policy-as-code sync) is in early access. Join the waitlist for access.
How do I upgrade the plugin?
Run openclaw plugins update zedly-shield to pull the latest version from npm. New protection features (like credit card redaction) are enabled by default on upgrade, so you do not need to edit your openclaw.json configuration.

Get Early Access to Zedly Shield

Join the waitlist for the cloud dashboard, fleet management, and policy-as-code. The open-source plugin is available now via npm.

You're on the list.

We'll reach out when early access opens. In the meantime, install the plugin and start protecting your OpenClaw deployment locally.

No spam. We'll email you when early access opens.

Protect Your Agentic AI Workflows Today

Install the plugin for free. Join the waitlist for the cloud dashboard.