ClawHub
Back to skill

Security audit

Incident Triage

Security checks for vulnerabilities and agentic risk

Overview

This incident-triage skill is a coherent troubleshooting workflow with disclosed GitHub helper scripts and no evidence of hidden, destructive, or persistence behavior.

Installers should treat this as an incident-response aid. Before relying on it during a real incident, fill in the runbook with your own approved service links and contacts, and review any incident issue body before running the GitHub issue creation script.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Hidden Instructions

High
Category
Prompt Injection
Content
# Runbook — Your Infrastructure Map

> ⚠️ **This is a template.** If you see `<!--` placeholder comments below, this file has not been filled in yet. Before using this skill in a real incident, populate this file with your actual services, endpoints, dashboards, and on-call contacts. Tell the user to fill it in now if it's still empty.

Fill this in with your services, endpoints, and tools. The triage framework references
this during Step 4 (Investigate) to know where to look.
Confidence
70% confidence
Finding
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The 'Use when' list includes very broad triggers such as 'something is down' and 'investigating an incident', which are common phrases that could match many ordinary troubleshooting conversations. The file does not provide constraints, explicit trigger phrases, or negative examples to clarify when this skill should activate versus when a different workflow should be used.

External Transmission

Medium
Category
Data Exfiltration
Content
| Service | Health Endpoint | Dashboard | Log Location |
|---------|----------------|-----------|-------------|
| <!-- api --> | <!-- https://api.example.com/health --> | <!-- Datadog/Grafana link --> | <!-- CloudWatch log group or log tool --> |
| <!-- web --> | <!-- https://app.example.com --> | <!-- link --> | <!-- link --> |
| <!-- worker --> | <!-- N/A (background) --> | <!-- link --> | <!-- link --> |
Confidence
50% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
| Component | CI/CD | Deploy Command | Rollback |
|-----------|-------|---------------|----------|
| <!-- API --> | <!-- GitHub Actions --> | <!-- Merges to main auto-deploy --> | <!-- Revert PR + merge --> |
| <!-- Web --> | <!-- GitHub Actions --> | <!-- Same --> | <!-- Same --> |

## On-Call
Confidence
85% confidence
Finding
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Static analysis

No suspicious patterns detected.