Back to skill

Security audit

Meta Knowledge Base

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent local knowledge-base tool, but it promotes broad automatic capture and durable plaintext storage of conversations and files without enough consent, scoping, or retention controls.

Review this carefully before installing. Use it only for data you intentionally want retained in a local knowledge base, avoid enabling automatic conversation or folder capture without explicit consent from affected users, and do not point it at secrets, credentials, private mailboxes, or broad home directories. Install dependencies in an isolated environment and prefer pinned, reviewed package versions.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (2)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:52
Finding

Unpinned and Unused Third-Party Dependencies Expand the Supply-Chain Attack Surface

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
meta_knowledge.py:126
Finding

Potentially Sensitive Knowledge Is Persisted in Plaintext Without Explicit Restrictive Permissions

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill explicitly advertises automatic capture from conversations, documents, emails, visited pages, and monitored folders without any privacy notice, consent flow, retention limits, or data minimization guidance. In a knowledge-base skill, this is especially dangerous because the core function is long-term aggregation and retrieval, which can silently accumulate sensitive personal, corporate, or credential-bearing information.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The OpenClaw hook example persists derived content from user messages after each interaction without notifying the user that their chat content may be stored for future retrieval. Because this occurs automatically in a post-message hook, users may reasonably assume the conversation is ephemeral when it is actually being converted into durable knowledge records.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
82% confidence
Finding

The skill documents capabilities that read local files and ingest content from paths, URLs, emails, and watched folders, but it does not declare any explicit tool scope or permission boundaries. This creates a mismatch between advertised behavior and security controls, increasing the risk of over-broad file access or unauthorized persistence when the skill is deployed in an agent environment.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The example instructs automatic persistence of conversation-derived content into a knowledge base without defining consent, scope, sensitivity filtering, or minimization boundaries. Even though the stored content is a summarized form, it can still encode personal interests, confidential topics, or sensitive business context that may later be surfaced by search or Q&A features.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The add() path persists arbitrary user-provided content and metadata to disk via _save() without any consent prompt, visibility, retention control, or redaction step. In a self-building knowledge-base skill, this is especially risky because users may supply sensitive conversation content, secrets, or personal data that gets silently stored under a durable local path.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

add_from_file() accepts an arbitrary local filepath, reads its contents, and ingests them into the knowledge base metadata and persistent storage without warning or access restrictions. In the context of a knowledge-capture agent, this increases the chance of exfiltrating or durably storing sensitive local files such as SSH keys, tokens, configs, or private documents if the skill is invoked on attacker-influenced paths.

Content

No source excerpt is available for this finding.