ClawHub
Back to skill

Security audit

Deep Current

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed local research tracker that writes user-managed thread and report files, with no evidence of hidden network access, credential use, destructive behavior, or prompt hijacking.

Install only if you are comfortable with an agent maintaining local research state in deep-current/currents.json and creating dated markdown reports in deep-current-reports/. If you enable the nightly cron workflow, review the prompt and timeout because it will let your agent search the web and write reports on a schedule.

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
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding

The skill describes and relies on local file reads/writes via a Python CLI and report generation, but it does not explicitly declare a tool scope such as permissions or allowed-tools. That mismatch can cause an agent or reviewer to underestimate the skill's filesystem capabilities, increasing the risk of unintended writes to local paths or broader-than-expected file access during autonomous runs like cron-triggered sessions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The markdown instructs users to schedule a job that writes findings to deep-current-reports/YYYY-MM-DD.md, which affects local user data. Although the README later notes where data lives, it does not clearly warn at the point of use that the skill will create and persist files on disk.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
94% confidence
Finding

The module docstring presents the available commands as the authoritative CLI interface, but it omits decay and covered even though both are registered in COMMANDS and executable. This is an active documentation/code divergence because the help text understates what the script can actually do.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The manifest and top-level docstring describe a local thread-management CLI for tracking topics, notes, sources, findings, and generating a digest. However, the code also implements decay and covered commands that automatically pause stale threads and scan report markdown files to extract covered topics/URLs, expanding behavior beyond the described thread-management scope.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.