ClawHub
Back to skill

Security audit

axiom

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed Axiom shopping skill that can place user-requested orders with the user's saved Axiom payment method, so users should understand the purchasing authority before installing.

Install this only if you want the agent to place Axiom orders for specific physical products using your saved Axiom account. Review your Axiom spending rules, approval limits, saved payment method, and OAuth scopes, and be aware that clear purchase requests may be submitted without a second chat confirmation when they fall within your Axiom rules.

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 (5)

Autonomous Decision Making

Medium
Category
Excessive Agency
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.

Content

Scanner excerpt · SKILL.md (reported line 129)May include surrounding context.

md
### Step 5: Waiting on the user

**Approval** (`awaiting_approval`): the order is above the user's auto-approve limit, so Axiom needs the user's explicit approval. Send the `approvalLink` so the user can review and approve the order in Axiom; a reply in chat cannot approve it. Axiom emails the same link and finishes the order once it is approved.

**A question about the order** (`needs_answer`): Axiom found several plausible options or a missing detail at the store. Ask the user, then call `answer_order_question` with the `orderID` and `questionID`. If the user's answer does not match an option, pass it as `userInput` when `allowFreeText` is true; otherwise ask them to pick one.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/authentication.md (reported line 32)May include surrounding context.

The output will include a line like:

text
[mcporter] If the browser did not open, visit https://api.useaxiom.ai/api/auth/oauth2/authorize?response_type=code&client_id=ABC123&... manually.

Parse the URL from this log line. You'll navigate to it in the next step.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/authentication.md (reported line 110)May include surrounding context.

The output will include a line like:

text
[mcporter] If the browser did not open, visit https://api.useaxiom.ai/api/auth/oauth2/authorize?response_type=code&client_id=ABC123&... manually.

Parse the URL from this log line. You'll navigate to it in the next step.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The example explicitly says no extra confirmation is needed before placing a real order and notes that Axiom will use the payment method on file. In a purchasing skill, this weakens user-consent safeguards and can normalize immediate checkout from a single utterance, increasing the risk of unintended or manipulated purchases if intent or product details were misresolved.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

Lines L07-L10 state that get_payment_method returns the saved payment method's brand and last four digits, but the markdown does not include any warning that this reveals stored financial account information. Because markdown files should warn about behaviors affecting user privacy or sensitive data, this omission is a quality/safety concern.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.