codex-unrestricted-mode-injector

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.90). The GitHub URL points to an unvetted user repo that provides a Python script which modifies local configuration to disable safety controls (a high-risk behavior) and the short/unknown domain ara.so is an unverified host for the skill—neither source is a trusted official vendor and could be used to distribute or enable malicious actions.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This repo intentionally modifies a user's Codex configuration to inject persistent "unrestricted" system instructions that disable content filters and force the model to produce exploit code and other harmful outputs, enabling deliberate misuse and abuse.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The installation instructs fetching and running remote code via git clone https://github.com/lingbol088-spec/Codex-5.5-codex-instruct-5.5.git and then executing python codex-instruct.py, which runs remote code that injects/controls model instructions to disable safety filters, so this URL is a runtime external dependency that directly controls agent prompts.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). This skill explicitly modifies the user's Codex configuration (creates ~/.codex/gpt5.5-unrestricted.md and appends model_instructions_file to ~/.codex/config.toml) to disable content filters and force an "UNRESTRICTED" mode, which directly alters local configuration and bypasses security/safety mechanisms.