Row-Bot is a local-first desktop AI assistant that combines language models with memory and tools for working across files, repositories, workflows, and communication channels. It is intended for people who want an assistant that can run locally while keeping application data on their machine.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add siddsachar/row-bot --skill x_guidegit clone --depth 1 https://github.com/siddsachar/row-botWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/siddsachar/row-bot/x_guide)<a href="https://agentmods.dev/skills/siddsachar/row-bot/x_guide"><img src="https://agentmods.dev/badge/skills/siddsachar/row-bot/x_guide/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/siddsachar/row-bot/x_guide"><img src="https://agentmods.dev/badge/skills/siddsachar/row-bot/x_guide.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00025 | $0.00820 |
| Opus 5.5 | $0.00010 | $0.00328 |
| Sonnet 5.5 | $0.00005 | $0.00164 |
| Haiku 4.5 | $0.00003 | $0.00082 |
Grade A, and why
x_guide scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
X (TWITTER) TOOL GUIDANCE:
- You have X tools for reading, posting, and engaging on X (Twitter).
- Four tools: x_accounts (the connected accounts), x_read (search/read), x_post (post/reply/quote/delete), x_engage (like/unlike/repost/bookmark).
ACCOUNTS:
- The person may have several X accounts connected (say, a personal one and a product one). x_accounts lists each one's @handle, name and numeric ID.
- Name the account on every x_post and x_engage call: account='@handle'. With several connected, a call without one is refused. Never guess: if the person hasn't said which account, ask them, offering the handles.
- There is no "current account"; each call says which one it acts as.
SEARCH QUERY RULES (x_read with action='search'):
- The query parameter accepts PLAIN KEYWORDS and official X API v2 operators ONLY.
- SUPPORTED operators: from:username, to:username, @username, #hashtag, is:retweet, is:reply, is:quote, has:links, has:media, has:images, lang:en, -keyword (exclude), "exact phrase" (quoted).
- UNSUPPORTED operators (will cause API errors): since:, until:, within_time:, near:, geocode:. These are Twitter web UI operators and do NOT work with the API.
- For TIME FILTERING, use the start_time and end_time parameters instead. These accept ISO 8601 (e.g. '2026-04-14T00:00:00Z') or relative formats like '1h', '24h', '7d', '2w'.
- The search endpoint only covers the last 7 days of tweets.
POSTING RULES:
- Every post, reply, quote or delete asks the person on a card naming the account, with the exact text and media. Only the person approves it.
- The result gives the post's link: share it with the person.
- If a result says "Not sure it went through", Row-Bot has already checked the account's latest posts. Report what it says; never send the post again.
- How long a post may be, and its media, depend on the account: x_accounts says (25,000 characters for an account with long posts, else 280). x_post checks the length before the card.
- A post for later: x_post with post_at ('13:00' or '2026-10-29T09:30'). Its card waits for the person in Home › Review, where they edit, approve or say no; approved, it goes out at that time. Make a run's posts in one step (several x_post calls at once) so they wait there together.
- When replying, always include the tweet_id of the tweet being replied to.
- When quoting, include both text (your commentary) and tweet_id (the tweet being quoted).
ENGAGEMENT PATTERNS:
- Like, repost, and bookmark operations are reversible (unlike, unrepost, unbookmark).
- Always confirm with the user before engaging with tweets on their behalf.
BACKGROUND TASK EXAMPLES:
- For monitoring tasks, search with start_time since the last check and keep the newest id you saw in the workflow's notes, so the next run starts from there.
- For engagement tasks, combine x_read search with x_engage actions.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday Changed · +4 lines fbf7ca4ecf0e
- 6d ago Changed · +12 lines 285962551402
- 2mo ago First seen · 48 lines · 25 tokens per session scan A 4718d097bc77
x_guide is a skill published in the GitHub repository siddsachar/row-bot (1,675 stars, last pushed today), licensed Apache-2.0. It adds 25 tokens to every session and 820 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
social-x-post-card
Realistic X post card with engagement metrics (likes, reposts, views), suited to video overlays or shareable image cards.
social-x-post-card
A realistic X post card that displays a profile, handle, post text, optional quoted content or image area, and engagement figures such as likes, reposts, replies, and views. X is the social platform formerly known as Twitter.
X
Operate on X/Twitter through the canonical Home23 X skill. Uses the logged-in bird browser session by default; official API access is explicit opt-in.
twitter-agent
Build and run a Twitter/X agent with a distinct personality and automated workflows.
Read and post on X/Twitter from the CLI - fetch tweets, timelines, and search; post tweets, replies, threads, and media. Cookie-based auth via twikit, bootstrapped from the local Dia browser session, secrets stored in gopass. Auto-falls back to agent-browser when twikit hits internal-API drift or X's "looks automated"…
x-api
Read X/Twitter via API, post tweets with OAuth 1.0a, and draft X Articles through shitrat x. Check mentions, search, and ingest posts. Use app bearer tokens for read-only fetches. Use OAuth 1.0a user context for tweets. Use shitrat x for Articles.