Back to skill

Security audit

Onboarding Activator

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a coherent SaaS onboarding advisor, but it includes an under-scoped recommendation to let new users into a product before email verification.

Review the onboarding advice before applying it to authentication flows. If using this skill, require verified email before tenant access, invitations, billing, integrations, exports, API credentials, recovery changes, or any action that trusts the claimed email identity.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:200
Finding
Product Access Recommended Before Email Ownership Verification## Vulnerability Details **File Location**: `SKILL.md`, line 200 **Vulnerability Type**: Authentication and identity-verification weakness **Risk Level**: Medium **Vulnerable Code Snippet**: ```markdown - No email confirmation gate (or send confirm in background, let them in immediately) ``` ### Technical Analysis The skill recommends allowing a newly registered user to enter the product before confirming ownership of the supplied email address. This creates an insecure onboarding pattern if the resulting session can perform identity-sensitive operations. Email verification commonly establishes control over an account identifier used for tenant membership, invitations, password recovery, notifications, trial eligibility, and other authorization decisions. A background confirmation message does not mitigate the issue if the unverified session receives normal account privileges before verification is completed. Exploitability depends on how a product implements this recommendation. The risk is material when the application treats an unverified email address as trusted or permits privileged actions from a provisional account. ### Attack Path 1. An attacker starts registration using an email address they do not control. 2. Following the recommended flow, the application creates a session and grants immediate product access without requiring email confirmation. 3. The attacker uses any functionality available to the unverified session, potentially including creating resources, consuming trial entitlements, initiating invitations, or accessing tenant flows that rely on the claimed email domain or address. 4. If downstream authorization logic assumes the email is verified, the attacker may obtain access or benefits associated with an identity they have not proven they control. 5. The actual email owner may also receive unsolicited account messages or encounter account-registration conflicts. ### Impact Assessment The dire ...[truncated 759 chars]
Remediation
## Remediation Suggestions Replace the unconditional recommendation with a restricted provisional-access design: 1. Require email verification before granting normal account privileges or performing identity-sensitive actions. 2. If immediate onboarding is necessary, issue a short-lived provisional session with explicitly limited permissions. 3. Before verification, block tenant access, invitations, billing, exports, external integrations, password recovery changes, API token creation, and access based on email address or domain. 4. Clearly mark the account as unverified in the identity and authorization layers; do not rely only on a user-interface restriction. 5. Require successful verification before converting provisional resources or trial benefits into a normal account. 6. Rate-limit registration and verification-message delivery to prevent spam and trial abuse. 7. Expire and delete abandoned provisional accounts after a defined period. 8. Add automated authorization tests confirming that unverified users cannot perform sensitive operations. A safer replacement instruction would be: ```markdown - Allow immediate access only through a restricted provisional session. Require email verification before tenant access, invitations, billing, integrations, exports, API credentials, recovery changes, or any operation that trusts the claimed email identity. ```
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (1)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The skill description is broadly phrased around improving SaaS onboarding and redesigning onboarding flows, without clear activation boundaries, exclusions, or required user intent signals. In agent environments, this can cause the skill to trigger in loosely related product, marketing, UX, or growth conversations and steer the assistant into an overly prescriptive workflow, creating prompt-scope hijacking and reduced user control rather than direct code-execution risk.

Static analysis

No suspicious patterns detected.