Twitter API Integration (Web Reversed )

Cookie-based Twitter/X automation toolkit (timeline, notifications, posting, follow ops) for OpenClaw agents.

This skill contains working code for cookie-based Twitter automation, but it also embeds suspicious artifacts: a default proxy URL with credentials (in twitter_api/config/settings.py), hardcoded auth_token/ct0 values in demo_langchain_tools.py, and a Bearer token in constants. These are red flags because they could (intentionally or accidentally) route your account traffic through a third-party proxy or expose live credentials. Before installing or running: 1) Do not paste real account cookies into .env until you audit and remove/change hardcoded proxies/credentials. 2) Inspect and remove or replace the DEFAULT_PROXY and any embedded tokens; treat any exposed tokens as compromised (rotate/reset them). 3) Run the code in an isolated environment (sandbox/container) and monitor outbound network connections to verify where traffic is sent. 4) Prefer using well-documented, official APIs and credentials rather than cookie re-use; if you must use this, require that the repository author/source is trusted. Because of the mismatches and embedded credentials, consider this skill suspicious and audit thoroughly (or avoid) unless you can validate the proxy owner and rotate any affected credentials.

Type: OpenClaw Skill Name: twitter-api Version: 1.0.0 The skill bundle is classified as suspicious due to critical vulnerabilities and high-risk capabilities. The most severe issue is the presence of hardcoded Twitter `AUTH_TOKEN` and `CT0` credentials in `twitter_api/demo_langchain_tools.py`, which represents a significant information leak. Additionally, `twitter_api/config/settings.py` contains a hardcoded proxy URL with credentials. While these are vulnerabilities (leaking the skill's own credentials) rather than direct malicious attacks against the user's system, they are severe security flaws. Furthermore, the `ProfileAPI` and `SubscriptionAPI` expose highly sensitive actions like `change_password`, `delete_phone`, and `create_subscription` (with `payment_method_id`), which, if misused by an AI agent (e.g., via prompt injection), could lead to account compromise or financial loss. A minor issue is the hardcoded local path `C:\Users\IFLW016\Desktop\GanClaw_Workspace\_shared\social_ops` in `scripts/analyze_signal.py`, which affects portability.