Back to skill

Security audit

Ai Side Hustle Assistant

Security checks for vulnerabilities and agentic risk

Overview

This is a simple Chinese-language side-hustle advice skill with no executable code, though users should be careful with the token-based CLI example.

Install only if you want Chinese-language AI side-hustle guidance. If you use the ClawHub publishing example, avoid pasting real tokens directly into shell commands when a safer interactive login, stdin, or credential-store method is available, and rotate any token accidentally exposed.

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
references/platforms.md:10
Finding

Authentication Token Exposed Through Command-Line Arguments

Content
View full analysis

Vulnerability Details

File Location: references/platforms.md, line 10
Vulnerability Type: Authentication token exposure through process arguments and shell history
Risk Level: Medium

Complete Vulnerable Snippet:

sh
clawhub login --token <token>

Technical Analysis

The documented login command instructs users to provide an authentication token directly as a command-line argument. After substitution of the placeholder with a real token, the secret may be exposed through:

  • Shell command-history files
  • Process listings or process-monitoring tools while the command runs
  • Terminal session recordings
  • CI/CD and automation logs
  • Diagnostic or audit telemetry that records command arguments
  • Scripts into which users copy the example

The project contains no executable code that automatically collects or transmits this token. The weakness is nevertheless actionable because following the documented procedure may disclose a live credential to other local users, administrators, or systems that retain command output and invocation metadata.

Attack Path

  1. A user replaces <token> with a valid ClawHub authentication token and executes the documented command.
  2. The shell, operating system, terminal recorder, or automation environment records the command or its process arguments.
  3. An attacker with access to the affected history file, process metadata, or logs retrieves the plaintext token.
  4. The attacker submits the stolen token to ClawHub.
  5. If the token remains valid, the attacker performs actions allowed by its assigned account permissions.

Impact Assessment

Exploitation requires access to local process information, shell history, terminal recordings, or relevant automation logs. A recovered token may allow unauthorized access to the associated ClawHub account and any operations authorized by that token, potentially including publishing or otherwise managing ...[truncated 182 chars]

Remediation
View remediation

Remediation Suggestions

  • Replace the command-line token argument with the official CLI's interactive secret prompt, standard-input mechanism, protected credential store, or another method that does not expose the token in process arguments.
  • If the CLI supports environment-based authentication, document use of a short-lived, narrowly scoped environment variable while warning that environment variables may still be visible in some diagnostic and process-inspection contexts.
  • Explicitly instruct users not to paste tokens into shell history, committed scripts, shared terminal sessions, or CI/CD logs.
  • Recommend short-lived, least-privilege tokens and routine rotation.
  • Advise users to revoke and replace any token accidentally exposed.
  • Configure automation systems to mask secrets and disable command tracing around authentication operations.
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 (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger scope is overly broad because it activates on generic topics like 副业, 赚钱, 变现, and 被动收入, which are common in normal conversation and may cause the skill to run when the user did not intend financial coaching. Unintended invocation can override more suitable skills or inject advice in unrelated contexts, increasing the risk of misrouting, poor user experience, and potentially unsafe financial guidance without explicit user consent.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill content and required output format are entirely Chinese-centric without documenting any user preference check or opt-in, which can force a language/context mismatch. While not directly leading to code execution or data compromise, it can cause incorrect routing, reduced comprehension, and exclusion of users who did not request Chinese output.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The file is entirely written in Chinese and does not state that the content is intended specifically for Chinese-speaking users or offer any language alternative. Under the policy rule for natural-language violations, a fixed language presentation without user opt-in or documented locale justification can be considered a locale-policy issue.

Content

No source excerpt is available for this finding.