TRUNKS SCORE / SIGNAL ENGINE
TRUNKS SCORE / SIGNAL ENGINEThe career score that reads your trajectory
Influence and momentum, 0 to 100. Recomputed on account sync, written every UTC day.
- signals_in
- 47
- pillars_active
- 5/5
- history_days
- 30
- recompute_ms
- 42
- confidence
- 0.91
live.demo
0
Scale 0 to 100. Emerging, rising, or dominant.
< 40
Emerging
40-69
Rising
70+
Dominant
live.telemetry
Live telemetry
Curve, pillars, radar, and platform sparks. Illustrative demo, not a real artist.
30-day evolution
Each UTC day adds a point. The score tracks trajectory.
Pillar mix
Illustrative. Real weighting adapts to available signals.
Signal radar
Five axes, one pulse. Relative read of active pillars.
Platform sparks
Monochrome silhouettes. Inbound momentum per connector.
Compute pipeline
Trunks does not scrape the open web. It records what arrives, extracts features, scores, then acts.
signal_engine.v1
- 01oauth_snapshot
Connect
OAuth Spotify, YouTube, TikTok, Deezer, Instagram. Your accounts, your truth.
- 02inbound.write
Ingest
Each sync writes an inbound snapshot. No nightly scrape of the catalog.
- 03renormalize(w)
Feature
Growth, engagement, market signals. Missing axes renormalize.
- 04trajectory_model
Score
Pillars fuse. Trajectory model activates when history is deep enough.
- 05utc_history.write
Act
Live score on you. UTC point in trunks_score_history. Trunks talks about it.
compute.tape
Engine session (demo)
- 01BOOT signal_engine v1 · locale=en · session=demo
- 02
Five pillars, one number
Each pillar feeds the Score. Machine tags visible, calibration internal.
Momentum
dyn.weightGrowth over 7 and 30 days, plus acceleration. Still climbing, or slowing?
- growth_7d
- growth_30d
- acceleration
Audience
dyn.weightEngagement versus volume. A reacting audience weighs more than idle stock.
- engagement_ratio
- active_listeners
- retention
Cultural signal
dyn.weightEditorial playlists, radio, recognition. The market sees you, or not yet.
- playlist_adds
- radio_hits
- recognition
Prediction
dyn.weightWhen history is deep enough, the model estimates if audience keeps rising. Otherwise the pillar drops out.
- trajectory_model
- mlPrediction
- history_depth
TrunksMag votes
dyn.weightA vote can lift. Absence does not lower. Community weighs without punishing silence.
- trunksmag_votes
- community_lift
- no_penalty
What the model refuses
Three anti-patterns. Trunks reads trajectory, not vanity.
reject
Follower vanity
A big isolated number is not a score. Trajectory counts, not the pose.
reject
Absence penalty
No TrunksMag yet? No drop. Lift only exists when there is a signal.
reject
Mass scrape
No nightly pull of everyone. We record what arrives: accounts, votes, prediction.
Methodology FAQ
Why does my score move?
Because inbound snapshots change: growth, engagement, market signals, or a prediction that activates. Each recompute also writes a daily history point.
How do I raise it?
Connect more platforms, grow for real, earn cultural signals, and let TrunksMag carry you. Ask Trunks for the next action, not just the number.
Is my data private?
OAuth tokens never enter the prompt. Agents read account tools. You talk to Trunks, not a public dashboard of your secrets.
How is the methodology built?
The Score combines momentum, audience, cultural signal, prediction, and TrunksMag votes. Calibration stays internal, like a credit score. You see the pillars, the curve, and the action.
Ask for your Trunks Score
Connect your accounts. Agents explain the number and what to do next.