The difference
Most AI writes into thin air.
RunSignl writes from your data.
RunSignl writes from your data.
It reads your real account — all-time top posts, real comments, real competitors — and turns what already works into what you make next. Every idea, hook and plan is grounded in evidence, not guessed. That's the whole game.
Provenance
Every line tells you
where it came from.
Most AI hands you a paragraph and hopes you don't ask where it got that. RunSignl stamps every block with its source — so you always know what's measured, what's calculated, and what's a draft waiting on you.
Real data
Scraped from your account
Your posts, your all-time top performers, real comments, the competitors you benchmark against. Pulled from the live platform — not remembered, not estimated.
e.g. your top post · real comment threads
Derived
Calculated from that data
Momentum, your winning pattern, gaps and recommendations. Arithmetic on real numbers — you can trace any of it back to the post it came from.
e.g. momentum · winning pattern
Drafted
Written by an agent, for you
Hooks, captions, scripts, replies. Clearly marked as written — never presented as fact, and never sent anywhere until you say so.
e.g. hooks · comment replies
When there's no data yet, RunSignl says so — "tracking begins after the next pull" — instead of inventing a number to fill the space. A dashboard that lies to you is worse than an empty one.
Inside RunSignl
Not a chat box.
A workspace.
Your account comes in as signal. The desk reads it, the board keeps what inspires you, and the workbench puts the whole team on a single idea. Every panel is stamped with where its content came from.
The Desk Derived
Momentum
Trending up
Winning pattern
Hook-first
▸ your best posts lead in the first 0.8s
Recommended next Drafted
01 Open on the payoff — not the setup REVIEW →
02 Answer your top comment on camera REVIEW →
03 Cut the intro from your best post REVIEW →
Inspiration board Real data
▸ Hooks
saved · breakdown ready
▸ Edits
saved · routed to Ideator
▸ Sounds
saved · queued
Breakdown Drafted
▚ the cut lands on the turn · caption enters late · loop closes on the payoff
Elevate · the idea workbench Drafted
▸ "Behind the scenes — how this one came together"
Ideator
Hook
Planner
Analyst
Community
Output Drafted · awaiting your yes
▸ hook, caption, shot list and a posting slot — all waiting on you APPROVE →
Inspiration
A save should start work.
Not rot in a folder.
You've got hundreds of saved reels you will never open again. RunSignl treats a save as the beginning of a job: it gets scored against a real baseline, broken down, and turned into a brief for your version — filed in a lane that decides which agent picks it up.
01
It lands
Scouted from the accounts you benchmark, or pasted in as a link — one or a hundred at a time.
Real data
02
It gets scored
Measured against that account's own average, so you know if it actually outperformed — or just looks good.
Derived
03
It gets broken down
The structure comes apart: where the hook lands, how the cut falls, when the caption enters.
Vision call
04
It becomes a brief
A plan for your version — built from the reel's structure and your own proven patterns.
Drafted
Tag it, and the work routes itself
The lane is the workflow. Filing a reel decides which agent picks it up and what happens next — you're not choosing a folder, you're assigning a job.
RecreateMake my version of this→ Ideator
StudyWhy did this work?→ Analyst
SoundsThe audio, not the video→ Hook & Script
ReferenceFiled for later — no work yet→ Planner
Friendly competition, on purpose
Add the accounts you measure yourself against. They do double duty — they keep the board full, and they give every reel a baseline to be scored against.
ScoutedYour benchmark accounts feed the board. Their real posts arrive on their own, so it's never empty on day one.
AddedPaste any reel link. It's pulled with real post data — caption, performance, the account behind it.
BaselineAdd a competitor and the next sync scores against them. Without one, RunSignl says so instead of inventing a number.
What it won't do: log into your account to read your private saves. That needs an unofficial login that risks a ban — so reels come in the two honest ways above instead.
The compounding loop
One idea. The whole team.
The winners compound.
RunSignl doesn't just make content — it learns from what lands and from every yes/no you give it.
STEP 01
Bring a signal
Your account, a mission, or a single idea. The engine takes it from there.
STEP 02
The team elevates it
Five agents develop it against your real numbers — ideas, hooks, a plan, the reply flow.
STEP 03
Keep the winners
You approve what ships. The loop learns your taste and your audience, and compounds.
Safety & control
Nothing leaves
without you.
RunSignl drafts. You decide. Every reply, every caption, every plan waits for a human before it goes anywhere — and that isn't a switch you can flip. It's the shape of the system.
Drafts only. Always.
The Community Manager writes replies. It does not send them. A human reads and sends every message — there is no auto-send switch to find.
You approve what ships
Every idea, hook and plan waits for your yes. The system learns from that yes — your taste is the product, not ours.
No personal details, ever
Nothing drafted carries personal details — location, contact info, schedule. Not a filter bolted on afterwards; a rule the agents are built under.
Your workspace is yours
Every account is isolated. Your data trains your agents and nobody else's — it is never pooled with other creators.
Who it's for
One operator.
Any size room.
Whether you run one account or a roster, the loop is identical: real data in, drafts out, you decide.
01 · Creator
You make the thing
You'd rather be in the studio than in the analytics tab. RunSignl reads the account, finds the pattern, and hands you the next three ideas already written in your voice.
02 · Manager
You run the account
You need the plan and the approvals in one place — and you need to know what's real. Every recommendation traces back to a post you can point at.
03 · Small team
You run several
Each account gets its own workspace, its own voice, its own agents. Switch between them without the crossover that ruins a brand's tone.
Questions
The straight answers.
01Is this just another AI caption writer?
No — and the difference is the input, not the model. A caption writer starts from your prompt. RunSignl starts from your account: your all-time top posts, your real comments, the competitors you benchmark against. It writes from evidence you can point at, which is why it can tell you why it suggested something.
02Does it post for me?
No. RunSignl produces drafts and a plan; you publish. DM and comment replies are drafted and always wait for a human to read and send. This is deliberate and it isn't configurable.
03What data does it actually read?
Public data from the platform: your posts and their performance all-time, public comments, and the public accounts you choose to benchmark against. It never logs into your account and never reads your private inbox or saved collections — that would risk your account, and we won't build it.
04Which platforms are supported?
Instagram is live. TikTok and YouTube are next — the platform switcher and the agents are already built for them; each just needs its own data connection. We'd rather ship one platform that genuinely works than three that half do.
05Who owns what it makes?
You do. Your data is yours, the drafts are yours, and your workspace is isolated — what RunSignl learns from your account is used for your account only, never pooled with anyone else's.
06Why is access limited right now?
RunSignl is in private beta. We're onboarding a small number of creators by hand so we can watch the loop work on real accounts and fix what doesn't. When it holds up, we open it up.
Private beta · onboarding by hand
Put your signal in motion.
We're taking on a small group of creators at a time, so we can watch the loop work on your real account. Tell us where to find you.
Instagram live · TikTok and YouTube next