Back to skill

Security audit

Brain Evolution Skill

Security checks for vulnerabilities and agentic risk

Overview

The skill is a memory/cache design, but it may persist user prompts and responses while making inconsistent privacy claims.

Install only if you are comfortable with agent interactions being stored locally for reuse. Before using it with real users or sensitive work, require explicit opt-in, secret/PII redaction before storage, per-user or per-session isolation, retention and deletion controls, and a clear way to disable or inspect persistent memory.

Vulnerability Patterns
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • 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
  • 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 (2)

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:48
Finding

Persistent Recording of Potentially Sensitive Agent Interactions Without Implemented Filtering

Content
View full analysis
Remediation
View remediation

T02 · Agent Memory Poisoning

Warning
Location
SKILL.md:244
Finding

Persistent Cache and Reflex Learning Can Propagate Poisoned or Stale Agent Responses

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The title and introductory sections are written in Chinese, and the file does not indicate that another language is available or that the skill is intentionally limited to Chinese-speaking users. This can violate language/locale policy when a skill imposes a specific language without user opt-in.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill claims '敏感对话不会被记录(可配置过滤)', but the rest of the documentation explicitly describes recording learning events, access traces, caches, and persistent storage without showing any implemented privacy filter. This creates a misleading privacy guarantee that may cause operators to deploy the skill under false assumptions, resulting in sensitive user data being logged to JSON/SQLite storage.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The file states that sensitive conversations are not recorded, but later lists sensitive-word filtering as a future improvement, directly contradicting the claimed current privacy protection. This inconsistency strongly suggests the protection is not implemented, increasing the risk of collecting confidential prompts, responses, or tool outputs into long-term storage.

Content

No source excerpt is available for this finding.