Back to skill

Security audit

Wiza

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent Wiza connector, but its setup instructions ask users to run downloaded installer scripts directly in a shell, which warrants review before installation.

Review the CLI installation path before using this skill. Prefer installing oo from a verified official package or a separately downloaded, inspectable installer with integrity checks, and confirm any Wiza write action before it runs.

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

Note
Location
SKILL.md:58
Finding
Unverified Remote Installer Retrieval and Immediate Execution## Vulnerability Details **File Location**: `SKILL.md`, lines 58–64 **Vulnerability Type**: Remote payload retrieval and execution **Risk Level**: Malicious ### 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 installation instructions retrieve mutable scripts from an external server and immediately pass them to `bash` or PowerShell's `Invoke-Expression`. The downloaded content is not pinned to an immutable release, saved for inspection, or verified using a cryptographic checksum or signature. Consequently, the code that ultimately executes can differ from what was available when the Skill was audited. Although the URLs use HTTPS and belong to the provider named by the Skill, HTTPS alone does not establish the integrity of a mutable installer. Compromise of the domain, hosting infrastructure, distribution pipeline, TLS termination, or installer publication account could replace the expected installer with arbitrary code. Installation is presented as a fallback for when the `oo` command is unavailable, rather than an operation performed on every invocation. This reduces exposure frequency but does not mitigate the arbitrary-code-execution risk. It also exceeds the Skill's normal `Bash(oo *)` least-privilege boundary because the installer itself can execute unrestricted shell operations. ### Attack Path 1. The Agent attempts to perform a Wiza operation and discovers that the `oo` CLI is unavailable. 2. It follows the documented first-time setup instructions. 3. `curl` or `Invoke-RestMethod` retrieves the current installer from the remote OOMOL endpoint. 4. The response body is passed directly to `bash` or `Invoke-Expression` without local integrity verification. 5. If the remote endpoint or software-distribution chain has been compromised, attacker-c ...[truncated 983 chars]
Remediation
## Remediation Suggestions 1. Remove direct `curl | bash` and `irm | iex` execution patterns. 2. Prefer a trusted platform package manager with a pinned package version and authenticated repository metadata. 3. If a standalone installer is required: - Pin an immutable, versioned release URL. - Download the installer to a local file without executing it. - Publish an expected SHA-256 or stronger checksum through an independent trusted channel. - Verify the checksum before execution. - Prefer a signed artifact and validate its signature against a documented publisher key. - Allow the user to inspect the downloaded script before running it. 4. Require explicit user approval before installing software or executing any downloaded script. 5. Run installation with the lowest privileges possible and avoid requesting administrator access unless a specific installation step requires it. 6. Document the exact files, directories, network endpoints, and privileges used by the installer. 7. Keep normal Skill operation restricted to the declared `oo` command boundary after verified installation.
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
Content
- **`oo: command not found`** — install the oo CLI (other platforms: <https://cli.oomol.com/install-guide.md>):

  ```bash
  curl -fsSL https://cli.oomol.com/install.sh | bash    # macOS / Linux
  ```

  ```powershell
Confidence
98% confidence
Finding
The skill instructs users to install software via a remote script piped directly into a shell, which prevents inspection before execution and creates a supply-chain execution path if the host, CDN, TLS termination, or distribution endpoint is compromised. In this skill context, the risk is heightened because the instruction is presented as an operational recovery step inside a trusted integration workflow, making users more likely to run it without scrutiny.

Vague Triggers

Medium
Confidence
97% confidence
Finding
The description says to use this skill for "ANY Wiza request" and "Whenever a task involves Wiza," which is an extremely broad activation condition for a markdown skill description. It does not provide narrower trigger boundaries, exclusions, or negative examples, so ordinary Wiza-related discussion could be interpreted as invoking the skill.

Static analysis

No suspicious patterns detected.