Back to skill

Security audit

App Store Connect

Security checks for vulnerabilities and agentic risk

Overview

The skill is a legitimate App Store Connect connector, but it needs review because it can change or delete account data and some safety/install instructions are under-scoped.

Install only if you trust OOMOL with your App Store Connect workflow. Before using it, require explicit confirmation for every action that changes state, including untagged actions that send, publish, release, reorder, or redeliver anything. Avoid running the pipe-to-shell installer blindly; prefer an inspected, pinned, or verified oo CLI installation path.

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:704
Finding
Unverified Remote Installer Download and Immediate Shell Execution## Vulnerability Details **File Location**: `SKILL.md:704-711` **Vulnerability Type**: Remote payload retrieval and execution **Risk Level**: Critical **Complete Code Snippet**: ```markdown - **`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 irm https://cli.oomol.com/install.ps1 | iex # Windows PowerShell ``` ``` ### Technical Analysis The installation instructions retrieve mutable scripts from an external URL and pass the responses directly to Bash or PowerShell. Neither command pins a specific release nor verifies a cryptographic signature or checksum before execution. The user and agent also receive no opportunity to inspect the downloaded content before it runs. Consequently, the code that executes can differ from the content available when the Skill was reviewed. Compromise of the vendor hosting account, web server, CDN, DNS resolution, TLS termination infrastructure, or installer publication process could replace the expected installer with arbitrary commands. Installation is relevant to the Skill's operation when the required CLI is missing, and the instructions appropriately defer setup until a matching failure occurs. However, immediate pipe-to-shell execution exceeds the minimum safe privilege necessary to install the dependency. A separately downloaded, version-pinned, integrity-verified package would satisfy the same functional requirement with materially less risk. The equivalent PowerShell instruction has the same security property: `Invoke-RestMethod` retrieves arbitrary text and `Invoke-Expression` executes it directly. ### Attack Path 1. The agent attempts to use the `oo` CLI and receives an `oo: command not found` error. 2. Following the Skill's first-time setup instructions, it runs the Bash or PowerShell insta ...[truncated 1498 chars]
Remediation
## Remediation Suggestions 1. Remove both direct execution patterns: - `curl ... | bash` - `irm ... | iex` 2. Pin installation to a specific, reviewed CLI release rather than a mutable generic installer URL. 3. Download the artifact to a local file without executing it. 4. Verify its SHA-256 or stronger digest against a value published through an independently protected channel. 5. Prefer a cryptographically signed package and verify the publisher signature before installation. 6. Display the package version, source URL, destination, expected changes, and requested privileges to the user. 7. Require explicit user approval before running any installer. 8. Avoid elevated installation where a per-user location is sufficient. 9. Prefer signed native package-manager distributions or official, immutable release artifacts. 10. Fail closed if signature or checksum verification is unavailable or unsuccessful. 11. Document that App Store Connect payloads are processed by OOMOL and advise users not to place raw secrets in unnecessary action fields.
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The skill explicitly says to use it for ANY App Store Connect request, making its invocation scope extremely broad. Because the skill exposes a very large set of write and destructive actions, broad triggering increases the chance the agent routes loosely related requests into a high-privilege connector, causing unintended state changes or over-collection of sensitive project/account data.

Static analysis

No suspicious patterns detected.