Social Media Agent
WarnAudited by ClawScan on May 10, 2026.
Overview
This skill is coherent, but it can use a logged-in browser and scheduled agent turns to publish X/Twitter activity without clear per-post approval or stop controls.
Install only if you are comfortable letting the agent use a logged-in X/Twitter browser session. Use a dedicated account or browser profile, require manual approval before every public post or engagement action, and avoid enabling cron schedules unless you have a clear way to review, pause, and remove them.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
The agent could publish, reply, quote, or otherwise engage from the user's account, creating reputational, business, or account-policy consequences.
The skill directs the agent to use browser automation to perform public, account-mutating actions on X/Twitter. The artifacts do not clearly require explicit user approval immediately before clicking the public Post button.
`browser` — Post tweets, engage with posts, take screenshots ... browser act → click Post button
Require explicit user confirmation for each post/reply/quote, define allowed action types, and keep high-impact browser actions manual by default.
If the browser is logged into the wrong or personal account, the agent may act publicly with that account's full session authority rather than a narrowly scoped permission.
Using browser automation instead of API keys likely relies on an authenticated browser session to act as the X/Twitter account. The artifacts do not clearly bound which account/session is used or what authority the agent receives.
Manage an X/Twitter account autonomously ... no API keys needed ... Ensure Chrome is running with remote debugging OR use OpenClaw's built-in browser
Use a dedicated browser profile or test account, confirm the logged-in handle before any action, and document the session/credential boundary clearly.
Scheduled posts could continue after setup and publish stale, unwanted, or unreviewed content from the user's account.
The skill explicitly instructs recurring scheduled autonomous agent turns for posting. The artifacts do not provide clear duration limits, kill-switch guidance, or mandatory review gates for scheduled posts.
Set up automated posting schedules ... Use `sessionTarget: "isolated"` with `payload.kind: "agentTurn"` for autonomous posting.
Make cron schedules opt-in, time-limited, easy to list and disable, and require approval before each scheduled public post.
Drafts or logs may persist across sessions and could be reused later, including if they become stale or contain information the user did not intend to keep.
The skill stores drafts, posting history, and engagement data in persistent memory/files. This is aligned with analytics tracking, but it can influence future posts and retain social strategy details.
`memory_search` / files — Track what was posted, engagement stats ... Save drafts in `memory/tweet-drafts-YYYY-MM-DD.json` ... Log posted tweets in `memory/social-log.json`
Review and periodically delete stored drafts/logs, avoid putting secrets or private information in drafts, and approve any memory-derived content before posting.