AI agents that ask before they act.
Intercepts every action your AI agent tries to take and checks it against a safety policy before execution. Local gating with browser dashboard, approval workflows, and audit ledger.
npx @authensor/safeclawCopyEvery tool call goes through a policy check before it runs.
Risky actions require your explicit approval. Nothing runs without your say-so.
Works with Claude and OpenAI out of the box.
100% open-source. MIT licensed. Free forever.
Action metadata only, e.g. "write /tmp/output.txt"
Your API keys, your files, your data. Always.
Setup wizard, task runner, and approval center in one UI.
Deny-by-default rules, visual editor, and dry-run simulation.
Append-only log with SHA-256 hash chain for every decision.
Badges on suspicious commands: obfuscation, credential access, exfiltration, destructive ops.
Spending limits, usage analytics, CSV/JSON export.
Swipe-to-approve from your phone. SMS and webhook notifications.
Sandboxed execution for untrusted workloads.
Fail-closed safety. Denies by default when disconnected.
Slack, Discord webhooks. Claude and OpenAI SDKs supported.
Your browser opens. A wizard walks you through everything.
npx @authensor/safeclawCopy