backbay-labsOfficial·10 Agent Skills Included

clawdstrike

Runtime security policies, threat detection, and signed audit trails

Enforces security policies on every file access, shell command, network call, and tool invocation before it runs. Blocks credential theft, secret leaks, prompt injection, and dangerous commands with fail-closed defaults. Signs every decision into a tamper-proof audit chain and helps investigate threats, tighten policies, and prove compliance.

285|33

↗

npx skills add backbay-labs/clawdstrike --all -g -y

Available:

Explains the policy engine architecture, built-in guards, and commands so the agent can evaluate actions, verify policies, and work within the fail-closed security model.

All Skills in This Repository (10)

Pure Emerald Level Indicators

📦 In Repo

backbay-labs

policy-guide

Explain ClawdStrike security policy configuration, guard management, and troubleshooting.

Threat Hunt

Automate security event investigation and threat hunting workflows.

security-review

Audits code changes for security risks and policy compliance in CI/CD.

hello

Generate personalized greetings and perform basic file operations in a secure environment.

bb-edr: Triage Skill

Convert ClawdStrike audit logs into incident reports with response plans.

hello-secure

Enforce clawdstrike policy checks on file access, network requests, and command execution.

build-scenario

Simulate security scenarios for the ClawdStrike policy engine.

security-audit

Audit ClawdStrike policy YAML files for vulnerabilities, compliance, and guard coverage.

observe-analyze

Analyze agent activity logs to synthesize and validate security policies.

tighten-policy

Validate and recommend ClawdStrike EDR security policy adjustments.

Official

Intermediate

Frequently Asked Questions

FAQPage Schema

How to install Clawdstrike?▼

Run `npx skills add backbay-labs/clawdstrike --all -g -y` in your terminal to install all skills in this suite globally.

How to stop AI agents from leaking secrets?▼

Clawdstrike checks every file write and network call against guards like SecretLeakGuard and EgressAllowlistGuard, blocking credential theft and unauthorized data exfiltration before it happens.

How to investigate suspicious agent activity?▼

The threat-hunt skill builds event timelines, correlates attack patterns, checks indicators of compromise, and generates incident reports mapped to MITRE ATT&CK techniques.

Can I create a least-privilege policy automatically?▼

Yes. The observe-analyze skill turns recorded agent activity logs into a candidate policy, which you can then test with scenarios and tighten guard by guard.

Does Clawdstrike work with Claude Code and Cursor?▼

Yes. It ships plugins for Claude Code, Cursor, and OpenClaw, plus SDKs for TypeScript, Python, Rust, and Go that enforce policies at the tool boundary.