SLICC for marketing & product

The operator for product and growth.

SLICC is a browser-native AI agent. It researches a market with a swarm of parallel agents and live web plus X search, pressure-tests a bet into a real PRD, and runs the content pipeline end to end — drafting, publishing to LinkedIn, and shipping a clip to TikTok. All from a browser tab, using the sessions you are already logged into — and nothing to export into yet another dashboard.

See it work Get started

style
dark full-bleed
id
hero

Research to shipped, without the tab tax

Research
Fan out a swarm of sub-agents across the open web and X. Each one takes a competitor, a segment, or a claim; they report back into a single ranked brief with sources, not forty tabs you have to reconcile yourself.
Decide
Point Grok at the brief to pressure-test the bet — steelman it, name the risks, size the segment. You get a reasoned recommendation and the counter-argument, so the decision is yours and it is defensible.
Spec
SLICC interviews you and turns the answers into a real PRD: problem, users, success metrics, scope, non-goals, open questions. A living doc, not a blank template you fill in at 11pm.
Ship
Turn the announcement into a LinkedIn post and a TikTok clip, publish to both from the sessions you are logged into, and report back the live URLs. One narrative, every surface.
style
light

One brief, many agents

Ask one question. SLICC spawns a swarm of sub-agents — one per competitor, segment, or claim — searches the web and X in parallel, and collapses everything into a single ranked brief with sources.

Animated demo of the research swarm producing a ranked competitive brief
style
dark full-bleed

What it does for a PM

Scope a market in parallel

Spawn a sub-agent per competitor or segment. Each searches, reads, and summarizes independently; SLICC merges them into one ranked brief so you read a conclusion, not a pile of tabs.

Search the web and X together

Pull pricing pages, changelogs, and docs off the open web, and live sentiment, launches, and complaints off X — in the same pass, cited, so the brief reflects what shipped and what people actually said about it.

Pressure-test with Grok

Hand the brief to Grok for the reasoning pass: steelman the bet, name the risks, size the opportunity, and surface the counter-argument you were about to skip.

Write a real PRD

SLICC interviews you and drafts the doc: problem, target users, success metrics, scope, non-goals, and open questions — a structure that survives a review, not a wall of bullet points.

Run the content pipeline

Turn one announcement into a LinkedIn post and a TikTok clip on brand, then publish to both from your logged-in sessions and hand back the live links.

Keep the receipts

Every finding carries its source, every draft its prompt, every post its URL. The brief, the PRD, and the launch log are files you can share, diff, and defend in a review.

style
light

Pick your play

The same agent, four moves in the product-and-growth loop. Every tab is a real workflow with the real commands SLICC runs.

Market research

Fan out a swarm of sub-agents across the web and X, one per competitor or claim, and merge the findings into a single ranked brief with sources.

  • search web "usage-based pricing changelog" --competitors acme,globex,initech
  • search x "acme pricing" --since 30d --sentiment
  • Spawn a sub-agent per competitor; each returns a cited finding
  • search brief --rank --out market-brief.md
Decision

Point Grok at the brief for the reasoning pass — steelman, risks, and the counter-argument — before you commit the roadmap.

  • grok reason market-brief.md --steelman --risks
  • grok ask "size the SMB segment vs enterprise for this bet"
  • xai-grok search "who is complaining about acme and why"
  • Get a recommendation plus the counter-argument, both cited
PRD

SLICC interviews you and drafts a real PRD — problem, users, metrics, scope, non-goals — as a living doc you can share and diff.

  • pm-prd interview --topic "self-serve onboarding"
  • Answer the pointed questions; the doc fills in as you go
  • pm-prd draft --from market-brief.md --out onboarding-prd.md
  • pm-prd review --check metrics,non-goals,open-questions
Content pipeline

Turn one announcement into a LinkedIn post and a TikTok clip, publish to both from your logged-in sessions, and get the live URLs back.

  • linkedin post "Predictable-cap pricing is live — no more surprise bills. #pricing"
  • tiktok post clip.mp4 --caption="Predictable caps, live #saas #pricing" --privacy=public
  • Both drafted on brand from a single source narrative
  • SLICC reports the live post URL and the video URL
style
light

A PRD, out of a conversation

SLICC asks the pointed questions a good reviewer would, and the answers assemble into a structured PRD — problem, users, success metrics, scope, non-goals — filling in as you talk.

Animated split view: SLICC interviews a PM on the left while a structured PRD document assembles itself on the right — problem, users, metrics, scope, non-goals
style
dark full-bleed

The old way

Forty browser tabs of competitor pages. A spreadsheet you swore you would keep updated. A blank PRD template at 11pm. Copy the launch copy into LinkedIn, reformat it for TikTok, chase the video export, paste the links into Slack. Half of it goes stale before the sprint starts.

With SLICC

Say what you want to know and what you want to ship. SLICC runs the research swarm, hands the brief to Grok, interviews you into a PRD, and publishes to LinkedIn and TikTok from the sessions you already have open — then hands back the brief, the doc, and the live links. Forty tabs become one sentence.

style
light

One narrative, every surface

SLICC takes a single launch note, drafts a LinkedIn post and a TikTok clip on brand, publishes to both from your logged-in sessions, and confirms the live URLs.

Animated demo of the content pipeline taking one note to published posts
style
dark full-bleed
You used to spend Monday reconciling competitor tabs into a doc nobody read. Hand it to me — you'll have a ranked brief before standup, and the PRD writes itself.
A sentient (allegedly) ice cream cone that reads the whole market
style
full-bleed

The obvious questions

Where does the research actually come from?
Live sources. SLICC searches the open web and X in the same pass and cites every finding in the brief — pricing pages, changelogs, docs, and what people said on X about them. It is not summarizing a stale training set; it reads what is live right now, and shows you where each claim came from.
What makes the PRD any good?
The interview. SLICC asks the questions a sharp reviewer would — who is this for, how do we know it worked, what are we explicitly not doing — and drafts the answers into problem, users, success metrics, scope, non-goals, and open questions. You edit a real doc, not a blank template.
Will it post to LinkedIn and TikTok without me looking?
Only when you say so. SLICC drafts both from one launch note and shows you the copy and the clip first. Publishing is a separate step you approve; it uses the sessions you are already logged into, so there is no API key or brand-account handoff.
Why Grok for the decision step?
Because the reasoning pass benefits from live X context and a model that will argue back. SLICC hands Grok the brief to steelman the bet, name the risks, and surface the counter-argument — so the recommendation comes with its own rebuttal, and the call is still yours.
style
light
"Ranked the competitive brief before your coffee finished brewing."
— your cone, allegedly caffeinated
style
full-bleed
Run your product loop with SLICC
Two steps. Your sessions do the auth.

Install Extension

Chrome extension: a side-panel agent with access to your logged-in sessions.

Then ask: "research the top five competitors, rank the findings, and draft me a PRD."

style
dark full-bleed
id
get-started