Grok Bot gives each Bot a name, a job, a conversation, and its own working computer. Skills capture the method behind a task, and routines decide when it runs. Felo skills are the research layer for that loop: live web answers, X signals, transcripts, and deck drafts your Bots can reuse without re-learning the work.
Save the process we just used as a skill called "Weekly market brief".
Include the sources, the output format, and the rule that
nothing gets posted without my approval.Grok Bot draws a clear line between three things: a Bot is the teammate with a job and a memory, a skill is a reusable set of instructions for doing one task, and a routine says when a Bot should run it. The recommended path is deliberate — do the task once, make it reliable, save the method as a skill, and only then automate it.
That structure rewards skills that produce a consistent result. If the output format changes every time, the routine that calls it produces a different artifact each run, and nobody can diff the results.
Felo skills are built for exactly that role: a fixed research request that returns findings with sources, a transcript, or a deck. Pair them with a Bot whose job is clearly defined — a Weekly Market Brief Bot, a Launch Watch Bot — and each run produces something comparable to the last.
Each pick maps to a Bot that would plausibly own it, and to a routine that would be worth scheduling.
Grok's home turf is X. This skill reaches posts, users, and replies directly, so a Launch Watch Bot works from what people actually posted.
Pull the top posts and replies about our launch today and summarize the recurring complaints.The default research step for any Bot that monitors a market, a policy, or a vendor. Sources make the routine output reviewable at a glance.
Research this week's changes for the accounts on my watch list and report only what is new, with sources.A Bot that runs a launch routine needs assets, not just text. SuperAgent drafts the logo or the product image, and the approval boundary still decides what goes out.
Create a launch image for the product update and hold it for my review before anything is posted.A Bot's shared computer and file handling make slides a natural output; this skill builds the deck from the research the Bot already gathered.
Build a ten-slide briefing from the research in this conversation and save it with the sources attached.Media watch is a common Bot job. Transcripts let a Bot report what was said in a video without you watching it.
Check this channel's new upload, pull the transcript, and tell me if it mentions our product.Grok Bot discovers packaged skills under Settings → Plugins and enables them per Bot. Pick the job you want covered: research, extraction, slides, transcripts, social monitoring, or a long-running workflow.
Add on Grok Bot
Live web research with sources you can cite
Settings → Plugins → Yours → enable felo-search for this BotTurn pages into structured, reusable data
Settings → Plugins → Yours → enable felo-web-fetch for this BotAnswers grounded in your own documents
Settings → Plugins → Yours → enable felo-livedoc for this BotTranscripts from talks, demos, and reviews
Settings → Plugins → Yours → enable felo-youtube-subtitling for this BotTrack posts and replies on X
Settings → Plugins → Yours → enable felo-x-search for this BotGenerate a deck from a single prompt
Settings → Plugins → Yours → enable felo-slides for this BotTurn notes and URLs into a deck outline
Settings → Plugins → Yours → enable felo-content-to-slides for this BotLong-running workflows that keep their memory
Settings → Plugins → Yours → enable felo-superagent for this BotCapture pages as snapshots and Markdown
Settings → Plugins → Yours → enable doc-screenshot-agent for this BotDraft X posts from your own material
Settings → Plugins → Yours → enable felo-twitter-writer for this BotGenerate a landing page from a brief
Settings → Plugins → Yours → enable felo-landingpage for this BotApple product research before you buy
Settings → Plugins → Yours → enable apple-buy-advisor for this BotThis list is for packaged skills. A skill your Bot learned from a run is saved from that run instead.
Written as the instruction a Bot owner would give, together with the skill that makes it repeatable.
Every Monday at 8:00, you want a short brief on three competitors, and you want to see what changed since last week.
Every Monday at 8:00, run the Weekly market brief skill against the current competitor list, flag only changes, and post the brief in this conversation.A dated brief with sources, plus a diff against the previous run so the summary gets shorter, not longer.
During a launch week, someone should watch X for complaints and draft responses, but nothing should be published automatically.
Every evening this week, check posts mentioning the launch, summarize the themes, and prepare drafts for approval. Never post without my confirmation.A daily themes summary and a queue of drafts that wait for a human yes.
A partner call or webinar recording lands in the conversation and needs to become a short deck for the team.
Pull the transcript from this recording, extract the commitments we made, and turn them into a five-slide recap.A recap deck grounded in the recording, with the commitments listed as action items.
A Bot is a name, a job, and a routine; skills are the part that makes the job repeatable without re-teaching it.
The first useful skill is the process you already ran by hand.
Routines can run on a schedule or wait for an event.
The Bot's shared computer and file handling mean the output can be more than text.
Grok Bot builds skills from work you have already done. This path gives it a clean first task so the saved method is worth reusing.
A job like 'Weekly Market Brief' produces reusable context. 'General helper' does not. Name the Bot, describe its ownership, and state the approval boundary in the description.
New chat → Create new agent
Name: Market Brief Bot
Description: Owns the weekly competitor brief. Reports only changes,
never contacts anyone, and requires approval before posting.Ask for something that needs live sources. This is the run the skill will be built from, so state the format you want to keep.
Research the three competitors on this list. For each, report pricing,
packaging, and launch news from the last seven days. Include a source
link and date for every item, and finish with a one-paragraph summary.Ask the Bot to record what it just did. A good skill names its inputs, its validation step, and its output format.
Save this process as a skill called "Weekly market brief". Include the
source list format, the change-detection rule, and the approval boundary.Skills are available across your Bots, but a Bot may still need the relevant plugin or login to use one. Enable the skill per Bot after saving.
Routines can run while the app is closed. Confirm the owning Bot, the schedule and time zone, the input source, and what happens when a source is unavailable.
Every weekday at 8:00 AM, run the Weekly market brief skill. Post the
linked watch list here. If a source is unavailable, report the failure
instead of using old data.Some skills arrive packaged rather than learned. Settings → Plugins is where supported connectors and packaged skills are discovered and installed, and where private skills are enabled for the current Bot.
Settings → Plugins → Yours
→ enable the skill for this Bot
→ then reference it with "/" in the desktop composerA Bot with a shared computer and a schedule can do real work while you are away. These habits keep its research useful and its actions contained.
Rules that should always hold belong in the description, not in a single conversation. State plainly that external contact, posting, and purchases require approval.
A routine is only reviewable if each claim carries a link. Put that requirement into the skill so every run inherits it.
Watch the first two executions before trusting the routine. Background work can run while your laptop is closed, so errors can accumulate quietly.
A shared Bot link exposes its identity, description, skills, and routines. Strip keys, internal URLs, and customer data before sharing one.
Run one research task, save it as a skill, and put it on a schedule. The second run should take less of your attention than the first.
Felo AI is an independent skill publisher. Grok Bot is a SpaceXAI product.
Start with a skill, then compose the exact models, live data, knowledge, and output tools your Grok Bot workflow needs — the same key every Felo skill already uses.
Use cost-efficient LLM access when your agent needs custom reasoning or generation.
Give agents current web and social signals instead of relying on stale context.
Ground work in reusable knowledge, then turn it into shareable deliverables.
Extend the workflow from research to visual and office-ready results.