Felo X Search on Doubao WorkDoubao Work

The weekly report needs to know what people said, not only what happened

A scheduled task can collect posts all week and still tell you nothing, because volume is not a finding. Felo X Search returns what was posted along with the signal each post carries, so the report says what changed instead of what trended.

Built for Doubao WorkRuns on a scheduleRead-only
Get an API keySee the platform guide
Doubao Work Felo X Search on Doubao Work

felo x-search "our product launch" --days 7

Bundle
felo-skills/felo-x-search
Query
Brand, product, or market
Comes back as
Posts with a signal tag
Cadence
Weekly, or per launch
Doubao Work logo
This skill runs on

Doubao Work

ByteDance's Doubao Work · agent work platform

Read the Doubao Work guide

Doubao Work turns a request into a finished file and can run it on a schedule. A Felo skill decides what goes into that file and keeps the sources behind every conclusion.

Runs on
Personal, team, and enterprise builds
Skills arrive by
Skill panel import, or the Felo CLI
Automation
Scheduled tasks and computer use
What you get back

Posts grouped by what they claim, so a report can say something

A launch produces hundreds of posts and about four distinct findings. Sorting them into signals is what turns a feed into a paragraph somebody can act on — and what keeps one loud complaint from looking like a trend.

@launch_watch3 hours ago

The new tier is cheaper than the old one for teams under ten seats. Prices are on the page, screenshot attached.

Concrete change
@mira_kYesterday

Support told us the migration is manual. That was not in the announcement and it changed our timeline.

Friction
@devnull_opsYesterday

Working fine here, but we only touch the API, so that is a narrow sample.

Scope limit
@press_wire_bot2 days ago

Most of today's coverage is recycled from the announcement post itself.

Echo, not reporting

The signal tag is what makes the list reviewable: it separates a documented change from a complaint before either reaches the report.

Where it stops

Three reasons a social report misleads

All three are avoidable once they are stated, which is why they belong on the page rather than in a footnote.

A post is a lead, not evidence

One reproducible report outweighs a hundred complaints. The signal tag exists so the difference survives into the summary.

Volume is not severity

A loud thread about a cosmetic issue can bury a quiet one about billing. Sort by what the post claims, not by how often it was repeated.

A scheduled query has a rate budget

Set the cadence against the budget. A job polling every minute spends the day rate-limited and reports nothing useful.

Add it on Doubao Work

Import it for a live check, or drive the CLI from a scheduled task

Most teams end up using both routes: the import for questions asked in the moment, and the CLI inside the task that assembles the recurring report.

Path A

Import the skill bundle

The shorter route, when your build exposes a skill or connector panel. Personal, team, and enterprise builds differ here, so check what your own account shows before planning around it.

  1. Clone the repositorygit clone https://github.com/Felo-Inc/felo-skills.git. Every bundle is a folder with a SKILL.md, and you only need the one you are installing.
  2. Import the folderBring that folder in through the skill or connector panel so social monitoring inside a recurring report is available while the task runs, not only afterwards.
  3. Confirm it is callableAsk the agent which skills it can use before you build a scheduled task around one. An imported skill that never gets called is the usual failure here.
Path B

Drive the Felo CLI

The route that works on every build. Install it once, then either run it yourself during setup or let computer use invoke it as part of a task on your machine.

  1. Install and store the keynpm install -g felo-ai felo config set FELO_API_KEY your-api-key-here
  2. Run the command by hand oncefelo x-search "our product launch" --days 7
  3. Then fold it into the taskAsk the agent to call the CLI when it needs current information or a deliverable, so the same command runs inside the task instead of beside it.
In practice

Two prompts for a launch week

The first decides whether you have a problem; the second writes the paragraph.

Type thisSearch X for posts about our launch in the last seven days and classify each one: confirmed change, friction, scope limit, or echo.What comes back

A sorted set where a single loud complaint is visibly a single loud complaint.

Type thisFrom that list, write the customer-reaction section of the weekly report. Quote two posts and say what the rest only repeated.What comes back

A section grounded in quotable posts rather than a general impression of the week.

Check it worked

Three checks before the section goes into the report

Fabrication in a social summary looks like confidence, so the checks are about specific posts rather than the tone of the whole.

1

Open one post

Pick a quoted handle and find it on X. A post you cannot locate is the fastest signal that the set was assembled rather than collected.

2

Check the dates against the claim

Posts should be recent relative to the change you are reporting, not merely recent relative to today.

3

Look for the counter-example

Ask the task for the posts that contradict its own summary. A set with no counter-examples was probably filtered for agreement.

FAQ

What Doubao Work users ask about Felo X Search

Can it watch for mentions continuously?
The skill answers a query when a task runs. Continuous watching is a schedule you set, and a weekly or launch-week cadence is usually the right one.
Can it reply or post on our behalf?
No. It reads. Nothing about it can write to an account, which is what makes it safe inside an automated task.
How does this differ from a media-monitoring service?
A monitoring service delivers volume and a dashboard. This delivers a classified set you can paste into a report, and it runs inside the same task that produces the deliverable.
What about posts in other languages?
Ask for the summary in your working language. Quoted posts are worth keeping in the original, with a translation beside them if the reader needs one.

Put the market reaction inside the weekly report

One skill, one key, and a scheduled task that reports what changed instead of what trended.

More on Doubao Work

Every Felo skill for Doubao Work

Five skills written for Doubao Work. They follow the same pattern on this platform, so you can combine them in one session without re-explaining your setup.

Beyond ready-made skills

One API key. A full agent toolkit.

Start with a skill, then compose the exact models, live data, knowledge, and output tools your Doubao Work workflow needs — the same key every Felo skill already uses.

Models

Use cost-efficient LLM access when your agent needs custom reasoning or generation.

  • LLM API
  • Chat completions
  • Model access
Explore LLM API →

Live information

Give agents current web and social signals instead of relying on stale context.

  • Web Search
  • X Search
  • Web Fetch
Explore live data tools →

Knowledge and output

Ground work in reusable knowledge, then turn it into shareable deliverables.

  • LiveDoc
  • PPT generation
  • Mindmaps
Explore knowledge tools →

Creative tools

Extend the workflow from research to visual and office-ready results.

  • Image generation
  • Page tools
  • Office outputs
Explore platform tools →