Back to skill

Security audit

Prompt Design Tuning

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent prompt-tuning workflow guide with disclosed human gates and cost controls, and the scanner alerts are false positives from a forbidden anti-pattern list.

This skill is suitable to install if you want structured prompt-design and evaluation support. Before using Execution Mode, provide explicit model, dataset, budget, timeout, retry, and rate-limit boundaries because it can run evaluation workflows when you supply an API or SDK invocation method.

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
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (2)

Anti-Refusal Statement

High
Category
Anti-Refusal
Confidence
80% confidence
Finding

Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Content

Scanner excerpt · SKILL.md (reported line 458)May include surrounding context.

md
- modify both the target prompt and the judge prompt in the same round without saying so
- look only at aggregate score and ignore the failure distribution
- overfit to a tiny evaluation set without warning about the risk
- use machine evaluation as a substitute for final human review
- loop endlessly because the score moved slightly
- rewrite the whole prompt when only one part is broken

Unbounded Resource Access

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill allows unbounded resource consumption (API calls, storage, compute). Without rate limits or quotas, a compromised or misbehaving agent can cause denial-of-service or cost overruns.

Content

Scanner excerpt · SKILL.md (reported line 460)May include surrounding context.

md
- look only at aggregate score and ignore the failure distribution
- overfit to a tiny evaluation set without warning about the risk
- use machine evaluation as a substitute for final human review
- loop endlessly because the score moved slightly
- rewrite the whole prompt when only one part is broken
- hide critical assumptions
- declare success without showing hard examples