Installs into .claude/skills of the current project.
Are you the author of X Com?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/tomcounsell-x-com-ai)
---
name: x-com
description: "Read X timelines or DMs, draft posts, and carry out specifically authorized replies, publication, or engagement."
---
# X Com
Use the available browser surface or X connector and verify the signed-in account; do not assume a handle from the Valor persona. Reuse the relevant tab and read full posts or conversation threads before replying. Refresh UI state after every mutation.
Execute only the requested activity. Checking a timeline or inbox does not authorize posting, mass likes, follows, or DM replies. Where the user already asked to publish or reply, draft and execute within that scope without redundant approval.
Ground content in actual source material or recent work. Fit the requested format and the account's observed limits; do not assume paid features. Keep short posts focused on one concrete point. Run $de-slop then $authenticity-pass, with at most two failed revision cycles. Never manufacture metrics or personal experience.
Prefer stable observed roles, labels, or test IDs over hashed CSS classes. The same submit control may serve posts and replies: verify the composer context and target thread. Dismiss upsell dialogs only after observing them; re-read after submission. Return verified permalinks or message state and disclose any unfinished draft. Inspect state before retrying an uncertain publish.
