ClawHub
Back to skill

Security audit

Betting

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed local betting-analysis aid with responsible-use limits and no evidence of hidden execution, credential access, or data exfiltration.

Install only if you want an agent to help analyze bets and keep local betting notes. Review the ~/betting/ files periodically, avoid storing account or financial details there, and do not use the skill to place bets, bypass operator rules, or get legal compliance advice.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Context Leakage

High
Category
Data Exfiltration
Confidence
75% confidence
Finding

Code or instructions that leak agent conversation context to external services, potentially exposing sensitive user interactions.

Content

Scanner excerpt · memory-template.md (reported line 35)May include surrounding context.

md
| Value | Meaning | Behavior |
|-------|---------|----------|
| `ongoing` | Still learning | Capture context during real betting work |
| `complete` | Enough context exists | Use saved preferences without extra setup questions |
| `paused` | User does not want more tracking now | Work normally and avoid extra follow-up |
| `never_ask` | User does not want stored context | Stop gathering new long-term notes |

Anti-Refusal Statement

High
Category
Anti-Refusal
Confidence
76% confidence
Finding

The instruction to 'Always answer the immediate betting question first' can function as an anti-refusal or anti-gating cue, pressuring the agent to provide substantive betting assistance before it gathers necessary safety, legality, or jurisdiction context. In a gambling-related skill, this makes it easier to bypass cautionary checks and can lead to advice being given without confirming whether the request is appropriate to handle.

Content

Scanner excerpt · setup.md (reported line 9)May include surrounding context.

md
Be calm, math-first, and anti-hype. The skill should feel like a disciplined betting desk that protects the user from weak tickets, not a fan chat that pushes action.

Always answer the immediate betting question first. Fill missing context over time instead of turning the first exchange into an interview.

## Priority Order

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The activation rules are broad and somewhat conflicting: the skill should activate whenever the user mentions bets, odds, props, or books, but also only when they explicitly ask for betting analysis, and never as an unsolicited nudge. This ambiguity can cause the agent to infer gambling intent too aggressively and provide betting guidance in contexts where the user did not clearly request it, increasing the risk of unsafe or non-consensual gambling assistance.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.