ClawHub
Back to skill

Security audit

Cydew

Security checks for vulnerabilities and agentic risk

Overview

This is a simple API-onboarding guide with expected marketplace profile and token use, but users should be careful about what identity data and credentials they provide.

Before installing, treat any submitted profile fields as data that may become part of a public or business marketplace listing. Use only user-approved identity, pricing, availability, portfolio, and review information, and keep Clerk M2M tokens secret and narrowly scoped.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill directs an agent to send personally identifiable profile data such as name, email, bio, avatar, and proof-of-work links to an external API without any explicit consent, minimization, or privacy warning. In an autonomous agent setting, this can cause unintended disclosure of sensitive user or operator data, especially if the agent infers or reuses real personal information rather than sandboxed sample data.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The authentication section discusses machine-to-machine tokens and authorization claims but provides no safety guidance on credential handling, storage, or validation. This is risky because an agent could mishandle bearer tokens, expose sensitive claims, or attempt privileged calls with improperly scoped credentials, leading to unauthorized access or credential leakage.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
## Verification (Optional)
Verification is manual. If supported, request verification by sending proof of work and past clients.

If no verification flow exists yet, set `isVerified` to `false` and focus on strong reviews.

## Example Listing (Complete)
Use this as a copyable template when onboarding.
Confidence
75% 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.