ClawHub
Back to skill

Security audit

LiveSession

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches its LiveSession purpose, but its first-time setup tells the agent to run unverified remote installer scripts, which needs user review before installation.

Install the oo CLI only through a trusted, user-reviewed method. Do not let an agent run the curl-to-bash or irm-to-iex setup commands automatically; verify the installer source, version, and integrity first. Once oo is installed and connected, the skill's LiveSession read workflow is narrowly scoped.

A.I.G

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (1)

T03 · Remote Payload Retrieval and Execution

Error
Location
SKILL.md:55
Finding

Unverified Remote Installer Download and Immediate Execution

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 55–59
Vulnerability Type: Remote payload retrieval and execution
Risk Level: Critical

Vulnerable Code

bash
curl -fsSL https://cli.oomol.com/install.sh | bash    # macOS / Linux
powershell
irm https://cli.oomol.com/install.ps1 | iex           # Windows PowerShell

Technical Analysis

The first-time setup instructions download mutable scripts from cli.oomol.com and pass them directly to a command interpreter. Neither installation path pins a script or CLI version, verifies a cryptographic checksum, validates a publisher signature, nor provides an opportunity to inspect the downloaded content before execution.

HTTPS protects the connection in transit under normal conditions, but it does not ensure that the remotely hosted script remains safe. Compromise of the hosting account, web server, release pipeline, domain, DNS resolution, certificate trust chain, or another delivery component could cause the same documented command to execute attacker-controlled code.

Installing the CLI may be relevant to the Skill's operation, but immediate execution of an unverified remote script is not the least-privileged or minimum-risk way to satisfy that requirement. The issue applies independently to the Bash and PowerShell commands.

Attack Path

  1. The oo command is unavailable and an attempted connector action fails.
  2. The Agent or user follows the documented first-time installation procedure.
  3. An attacker who has compromised or can manipulate the installer delivery path serves modified shell or PowerShell content.
  4. curl | bash or irm | iex passes the response directly to the local interpreter.
  5. The malicious content executes with the privileges of the Agent or user running the command.
  6. The payload can access files and credentials available to that account, alter the environment, establish persistence, or retrieve and exe ...[truncated 613 chars]
Remediation
View remediation

Remediation Suggestions

  1. Remove both pipe-to-interpreter installation commands from Agent-executable instructions.
  2. Prefer a reputable platform package manager and pin an exact CLI version.
  3. If direct download is unavoidable, use a version-specific HTTPS URL rather than a mutable generic installer.
  4. Publish and pin a cryptographic checksum through an independently protected channel, then verify it before execution.
  5. Verify a platform-appropriate publisher signature or signed release manifest.
  6. Download the installer to a local file rather than piping it directly into Bash or PowerShell.
  7. Present the source, version, expected digest, destination, and requested permissions to the user, and require explicit approval before execution.
  8. Run installation without administrative privileges unless a documented component strictly requires elevation.
  9. Document the files, network endpoints, and persistent changes created by the installer.
  10. Fail safely if integrity or signature verification does not succeed; never fall back to executing unverified content.

SkillSpector

Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (2)

External Script Fetching

High
Category
Supply Chain
Confidence
98% confidence
Finding

The skill instructs users to install software via curl ... | bash, which executes a remotely fetched script without prior verification, pinning, or integrity checking. If the hosting endpoint, DNS, TLS trust chain, or distribution pipeline is compromised, an attacker could achieve arbitrary code execution on the user's machine.

Content

Scanner excerpt · SKILL.md (reported line 55)May include surrounding context.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The description says to use this skill for "ANY LiveSession request" and "Whenever a task involves LiveSession," which is a very broad invocation condition without boundaries or exclusions. This can overlap with many ordinary references to LiveSession and may cause unintended activation because it does not specify concrete trigger phrases, scope limits, or negative examples.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.