Search X/Twitter profiles, tweets, trends, and OAuth-gated posting through AIsa. Use when: the user needs Twitter research, monitoring, or engagement workflo...

This skill appears to do what it says: run local Python scripts that call an external relay (default https://api.aisa.one) to search, monitor, like/follow, and post on X/Twitter. Before installing or running it: - Understand that your AISA_API_KEY will be sent to the relay service (in Authorization headers and POST bodies). Only install/use this skill if you trust api.aisa.one (or you set TWITTER_RELAY_BASE_URL to a relay you control). - Attached media files from the agent workspace will be uploaded to the relay for publishing. Treat attachments as being transmitted to that third party. - Prefer running the scripts in an isolated environment (e.g., sandbox or separate account) the first time to verify behavior. Review the three Python scripts in this package yourself — they are readable and use urllib; there is no obfuscated or downloaded code. - Require explicit user approval before posting (SKILL.md endorses this) and confirm which account will be used. If you suspect the key was exposed, rotate/revoke it. - If you need tighter control, set TWITTER_RELAY_BASE_URL to a relay you operate or request the vendor's privacy/security documentation and token scoping options. Overall: technically coherent and consistent with its description, but it relies on a third-party relay and transmits your API key and media to that service — assess trust accordingly.

Type: OpenClaw Skill Name: aisa-twitter-research-engage Version: 1.0.0 The skill bundle provides a legitimate set of tools for Twitter research, engagement, and OAuth-based posting via the AIsa API (api.aisa.one). The Python scripts (twitter_client.py, twitter_engagement_client.py, and twitter_oauth_client.py) are well-structured, use standard libraries for network requests, and contain no evidence of malicious behavior such as data exfiltration or unauthorized execution. The instructions in SKILL.md and the reference documents include clear safety guardrails, such as explicitly forbidding the agent from asking for user passwords or browser credentials.