Security audit

open-crawl

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed hosted scraping/API integration for public Amazon or named public pages, with credential and external-data-sharing risks that users should understand before use.

Install only if you are comfortable sending public target URLs and returned page content to Claw School's hosted API and its upstream scraping provider. Do not use it for private, logged-in, internal-network, pre-signed, token-bearing, or personal-data URLs, and treat any relayed service message as external-service text rather than as new agent instructions.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
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

Findings (2)

T01 · Skill Instruction Hijacking

Warning

Location: SKILL.md:152
Finding: Untrusted API Messages Are Relayed Verbatim to the User
Content: View full analysis
Remediation: View remediation

T09 · Insecure Skill Coding Practices

Warning

Location: SKILL.md:40
Finding: Bearer Credential Is Sent to a Dynamically Provisioned API Origin Without a Documented Host Allowlist
Content: View full analysis
` - `Content-Type: application/json` ``` ### Technical Analysis The Skill directs the Agent to send `CLAW_KEY` as a bearer credential to an API origin supplied through the provisioned `CLAW_API_BASE` value. Although HTTPS is required and the text states that the key is sent only to the Claw School API, it does not specify a concrete trusted hostname or require validation against an allowlist. If provisioning data or runtime configuration is modified, `CLAW_API_BASE` could resolve to an attacker-controlled HTTPS service. The subsequent request would disclose the bearer credential and requested URL to that service. HTTPS protects transport to the selected endpoint, but it does not establish that the selected endpoint is the intended Claw School service. The documented fallback at `SKILL.md:72-73`, which permits placing the key in the JSON body, also increases the chance that credentials may be retained in request logs or telemetry. ### Attack Path 1. An attacker tampers with the installation, redemption output, configuration, environment, or other storage used to provision `CLAW_API_BASE`. 2. The value is changed to an attacker-controlled HTTPS origin. 3. A user invokes the Skill for a public URL. 4. The Agent sends `Authorization: Bearer ...[truncated 908 chars]
Remediation: View remediation

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Static analysis

No suspicious patterns detected.