Back to skill

Security audit

Fluxguard

Security checks for vulnerabilities and agentic risk

Overview

The Fluxguard skill is mostly purpose-aligned, but its setup instructions tell an agent to run unverified remote installer scripts, which creates avoidable local code-execution risk.

Install only if you are comfortable with OOMOL mediating Fluxguard actions and with reviewing every write or delete request before approval. Do not let an agent run the listed curl-to-bash or irm-to-iex installer automatically; use a safer verified installation path for the oo CLI when available.

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:66
Finding
Unverified Remote Installer Scripts Executed Directly by Shells## Vulnerability Details **File Location**: `SKILL.md`, lines 66–70 **Vulnerability Type**: Remote payload retrieval and execution **Risk Level**: Critical ```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 setup instructions pipe responses from mutable external URLs directly into Bash or PowerShell. The remote content is executed without version pinning, checksum validation, digital-signature verification, or prior inspection. HTTPS protects the connection in transit but does not establish that the retrieved script is immutable or safe. If the hosting infrastructure, publisher account, DNS resolution, TLS trust chain, or installer itself is compromised, the external response can be replaced with arbitrary commands after this Skill has already passed review. Installing the CLI may be necessary for the declared Fluxguard functionality, but immediate execution of unverified network content is not the minimum-risk installation method. The Skill could instead direct users to a pinned, authenticated release artifact or a trusted package manager. No evidence establishes that the current installer is intentionally malicious; the vulnerability arises from its unsafe delivery and execution mechanism. ### Attack Path 1. The `oo` CLI is absent, causing the user or Agent to consult the first-time setup instructions. 2. An attacker compromises or gains control over an installer endpoint or another relevant component of its delivery chain. 3. The attacker modifies the response from `install.sh` or `install.ps1` to include arbitrary commands. 4. The user or Agent runs the documented command. 5. Bash or PowerShell executes the network response immediately, without integrity or authenticity validation. 6. The payload operates with all permissions available to the invoking process and ...[truncated 1061 chars]
Remediation
## Remediation Suggestions 1. Remove both direct pipe-to-shell installation commands. 2. Publish installation artifacts under immutable, version-specific release URLs. 3. Provide expected SHA-256 or stronger hashes through an independently authenticated release channel, and require users to verify them before execution. 4. Digitally sign release artifacts and document mandatory signature verification against a pinned publisher key. 5. Prefer a reputable package manager that supports version pinning, package signing, and reproducible provenance. 6. If script installation remains necessary, require separate download, verification, inspection, and execution steps rather than passing the network response directly to a shell. 7. Document that installation must run without administrative privileges unless a specific operation strictly requires elevation. 8. Pin the installer or CLI version so later server-side changes cannot silently alter the payload reviewed for this Skill. 9. Maintain release provenance, a software bill of materials, and a documented key-rotation and incident-response process for installer compromise.
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

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
97% confidence
Finding
The skill instructs the agent to install software by piping a remotely fetched script directly into a shell (`curl ... | bash`). This creates a supply-chain and remote code execution risk: if the install endpoint, network path, or hosting account is compromised, arbitrary code will run immediately on the user's machine without review. In this skill context, the danger is heightened because the install step is presented as an operational fallback inside an agent workflow, making unsafe execution more likely.

Static analysis

No suspicious patterns detected.