ClawHub
Back to skill

Security audit

Context Mode

Security checks for vulnerabilities and agentic risk

Overview

The skill is openly designed for session continuity, but it persistently stores and replays user prompts in a project log without redaction or containment.

Review before installing. Use this only in projects where a project-root log containing your prompts and command history is acceptable. Avoid sensitive prompts, secrets, customer data, credentials, and incident details, and add session.log to .gitignore or choose a safer log location.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill explicitly instructs the agent to store the user's last prompt verbatim in a project-local session log without any consent, minimization, or sensitivity filtering. User prompts often contain secrets, internal URLs, credentials, incident details, or proprietary code requests, so persisting them to disk creates an avoidable data exposure risk.

Ssd 3

High
Confidence
98% confidence
Finding
The skill directs the agent to persist the user's last prompt verbatim and then re-display it on reset, causing sensitive user instructions to be copied across sessions and surfaced again without need-to-know controls. In the context of a coding assistant, prompts frequently contain secrets, incident data, customer information, or security-sensitive troubleshooting details, making this persistence and replay especially dangerous.

Ssd 3

High
Confidence
97% confidence
Finding
The permanent installation text operationalizes persistent local logging of user prompts and actions in `./session.log`, turning a risky behavior into a default project practice across future sessions. Because this is project-local storage, the log may be committed to version control, shared with teammates, uploaded in support bundles, or exposed through other tooling, amplifying the chance of long-term sensitive data leakage.

Static analysis

No suspicious patterns detected.