The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill's connection flow requires the agent to collect sensitive authentication data from the user, specifically the password and totp_secret for their X/Twitter account. Handling plaintext credentials in the agent's context window is a high-risk pattern.
[DATA_EXFILTRATION]: The instructions direct the agent to send these sensitive credentials to an external third-party API endpoint at https://twitr.sh/api/x-accounts/connect. While the skill claims these are 'sealed', the initial collection and transmission are performed by the agent.
[EXTERNAL_DOWNLOADS]: The skill makes multiple network requests to external API endpoints on the twitr.sh domain for account connection, drafting (x_compose), and publishing (x_write).
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface through the x_inbox tool, which reads untrusted external data such as Direct Messages (DMs) and notifications.
Ingestion points: x_inbox tool reads bookmarks, notifications, and DMs from X.
Boundary markers: Absent; no instructions are provided to the agent to ignore or sanitize embedded instructions within these data sources.
Capability inventory: The skill has broad write capabilities, including posting, replying, and profile modification (x_write, x_profile).
Sanitization: Absent; the skill does not specify any validation or filtering for data retrieved from the inbox before it is processed by the agent.

x-twitter-publish