ClawHub
Back to skill

Security audit

FormPass Submit

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward FormPass form-submission helper, but users should confirm before sending personal or sensitive form data.

Install only if you want agents to submit FormPass-enabled forms. Before posting, review the destination, fields, and any Agent ID token, and require explicit user approval for personal, financial, regulated, or otherwise sensitive data.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The skill’s invocation guidance is broad enough to trigger on generic requests to fill out or submit forms, which can cause an agent to use this capability in situations where the user did not explicitly consent to sending data to a third-party service. Because the skill ultimately performs real external submissions, over-broad routing increases the risk of unintended disclosure of personal or sensitive information.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The submission instructions describe how to POST form data and an agent identifier to FormPass, but they do not prominently warn that this transmits user-provided content to an external service. This can mislead an agent or user into treating the action like local form-filling rather than outbound data sharing, increasing privacy and compliance risk.

External Transmission

Medium
Category
Data Exfiltration
Content
POST your data as JSON. Include your Agent ID as a Bearer token if you have one (this identifies you as a verified agent).

```bash
curl -s -X POST "https://form-pass.com/api/submit/FORM_ID" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_AGENT_ID" \
  -d '{
Confidence
92% confidence
Finding
This instruction performs an external POST of form contents and optionally an agent bearer token to a remote service. In the context of a form-submission skill, the transmission is expected, but it is still security-relevant because it can expose user data or credentials if invoked without explicit consent, sufficient validation, or trust checks on the destination.

External Transmission

Medium
Category
Data Exfiltration
Content
echo "$SCHEMA" | jq '.fields[] | {name, type, required}'

# 3. Build and submit your data
curl -s -X POST "$HOST/api/submit/$FORM_ID" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer fpagent_your_id_here" \
  -d '{
Confidence
92% confidence
Finding
The example shows a full end-to-end external submission flow that sends structured data and an authorization header to a remote host. Although aligned with the skill’s purpose, it materially enables data exfiltration to an external service if the skill is invoked too readily or used with sensitive user content.

Static analysis

No suspicious patterns detected.