Current Target

No scan yet

idle
0 info

Policy Decision

allow

Paste a skill or choose an example to scan.

Pre-Install Gate

Waiting for scan

Run a scan to see whether ClawGuard would install, pause, or block before files are trusted.

npx --package @denial-web/clawguard clawguard install ./skill --to ./.agents/skills --policy governed

Run Plan

Waiting for scan

Scan a skill, then generate the policy and model route before trusted install.

Skill Gate allow
Model Profile none
Budget not checked
npx --package @denial-web/clawguard clawguard run-plan --skill ./skill --task "Install and run this skill"

Approval Loop Demo

Prove the full guarded install loop

Create a harmless temporary skill, scan it, write an approval request, record an owner approval, and install only after the decision is applied.

install hook policy gate owner approval apply
npx --package @denial-web/clawguard clawguard approvals demo-flow --keep
npx --package @denial-web/clawguard clawguard openclaw install ./skill --to ./.agents/skills --approval-out ./.clawguard/approvals.jsonl --approval-mode always

Agent Dashboard

Local Runtime State

Approvals, memory, audit, and bridge state will appear here.

Pending 0
Audit 0
Memory 0
Bridge off

Approvals

No approvals loaded.

Audit

No audit events loaded.

Memory

No memory loaded.

Bridge

No bridge state loaded.

Business SOP Gate

Choose a SOP demo

Pick a business workflow to see how ClawGuard checks evidence, thresholds, approvals, and blocked actions before an agent marks work complete.

Missing Evidence 0
Approvals 0
Thresholds 0
Blocked Actions 0
npx --package @denial-web/clawguard clawguard sop init --industry cafe --out cafe-close.json
SOP results will appear here.
Critical0
High0
Medium0
Low0
Files 0
Workspace Skills 0
ClawHub Entries 0
Dependency Manifests 0

Findings

Scan results will appear here.