Back to skill

Security audit

Vibecoding Pro

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Generator/Evaluator QA workflow template, with some operational cautions around browser/API testing but no hidden persistence, exfiltration, or destructive behavior found.

Use this skill as a QA workflow template. Run browser/API evaluator loops only against localhost, staging, or disposable test systems with non-production credentials, and review any adapted generator/evaluator code before giving it authority to write code, call authenticated APIs, or submit forms.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Memory PoisoningPersistent Context Injection, Context Window Stuffing, Memory Manipulation
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (6)

Memory Manipulation

High
Category
Memory Poisoning
Confidence
80% confidence
Finding

Skill manipulates agent memory, state, or stored context. Memory corruption can alter personality, override safety rules, or cause unpredictable behavior.

Content

Scanner excerpt · references/architecture.md (reported line 216)May include surrounding context.

md
| No generator code | Prevents reasoning from implementation | Evaluator sees only deployed URL |
| Action-first | Evaluator uses artifact, then reads | Playwright interactions precede all analysis |
| Structured JSON only | Prevents sycophantic drift | No freeform "looks great!" output |
| Fresh context/round | Prevents anchoring to previous round | Clear context each iteration |

---

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
84% confidence
Finding

The skill describes and references executable workflow components (scripts/iteration_loop.py, scripts/calibrate_evaluator.py) and operational capabilities that imply file read/write behavior, but it declares no explicit tool scope or permissions boundary. That creates ambiguity about what the agent is allowed to access and can lead to overbroad file operations when the skill is installed or adapted into an agent runtime.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill instructs users to perform browser automation and direct HTTP/API testing against deployed artifacts, but it provides no safety guidance about using test environments, non-production credentials, rate limits, destructive actions, or mutation of real data. In practice, an evaluator loop that clicks through forms and calls endpoints can create, modify, or delete records or trigger side effects in external systems.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest describes a working skill that separates generation from independent QA through real browser interaction. In this file, both core operations are placeholders that immediately raise NotImplementedError, so the claimed workflow is not actually implemented here.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

The top-level docstring states that the script provides a complete framework and shows direct usage, which suggests runnable end-to-end behavior. However, the generator and evaluator functions later raise NotImplementedError, so the documentation overstates operational completeness.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
79% confidence
Finding

This code performs a file write operation when output_log is provided, but the implementation contains no confirmation prompt and no explicit warning in comments or docstrings that running the script will create or overwrite the target file. Because file writes are in scope for missing user warnings on code files, this merits a low-severity disclosure finding.

Content

No source excerpt is available for this finding.