One front door to Ghost Intelligence
Everything Pulse can already do, organised so you understand it in five minutes. Each capability answers three things: what is happening, why it matters, and what you can do next.
Executive Dashboard
Auto-refreshes every 20sSystems Online
—
checking…
Average Risk
—
checking…
Active Alerts
—
checking…
Data Provenance
—
checking…
Service
—
checking…
Ghost Console — Operate
PreviewType a goal and watch Ghost run the operating loop: classify intent → plan → recall data → score risk → stop at approval. Read-only preview; execution is disabled.
Intelligence
1 capabilitiesMemory
PartialRemembers context, facts and beliefs.
What: Stores and consolidates episodic and semantic memory, tracking beliefs and contradictions.
Why: Memory gives the system continuity instead of starting from zero each time.
Next: Open Memory recall. Durable propose/approve lives in Cognition Lab.
Markets
5 capabilitiesTrading Brain
WorkingWeighs market signals into a trade decision.
What: Fuses price action, macro and math signals into a gated BUY/WAIT/SELL call with rationale.
Why: Shows the decision engine applied to a real, high-stakes domain.
Next: Open Markets to watch the trading brain decide (simulated execution).
Market Signals
WorkingLive market data from Polygon, Finnhub and Tradier.
What: Streams price, breadth and volatility signals.
Why: Real market data is the raw input for trading intelligence.
Next: Open Markets to view the live market signals.
Weather Intelligence
WorkingLive weather and severe-weather alerts.
What: Pulls real conditions and alerts from OpenWeather and NWS.
Why: Weather drives demand, risk and infrastructure stress.
Next: Open the Weather view for live conditions.
Traffic Intelligence
WorkingLive traffic flow and incidents on a map.
What: Renders TomTom traffic flow, incidents and a routing brain.
Why: Geospatial movement is a leading indicator across many systems.
Next: Open the Operational Map to see live traffic.
Cross-Domain Risk
WorkingOne risk score across every domain.
What: Combines all live signals into a single risk level with anomalies and recommended actions.
Why: Cascading failures only appear when you look across domains.
Next: Open Core Health to see the unified risk picture.
Operations
12 capabilitiesProvider Health
WorkingLive status of every external integration.
What: Tests 13 providers for configuration, latency and errors.
Why: Intelligence is only as good as the feeds behind it.
Next: Open Core Health to see the live provider scoreboard.
Memory Health
PartialIs the memory system stable and reliable?
What: Reports memory reliability and consolidation state.
Why: Unstable memory quietly corrupts every downstream decision.
Next: Open Core Health to check memory reliability.
Agent Status
WorkingSix specialist agents and their health.
What: Shows planner, analyst, verifier, memory, operator and observer agents with live lifecycle.
Why: The system is a team of agents — you should see the team.
Next: Open Coordination to meet the agents.
Diagnostics
WorkingDeep system diagnostics in one place.
What: Aggregates per-subsystem diagnostics and smoke checks.
Why: When something is off, this is where you look first.
Next: Open Diagnostics for the full system read-out.
Mission Command
WorkingDefine and track missions end to end.
What: Manages mission lifecycle with a persistent mission store.
Why: Missions give the system goals to organise its work around.
Next: Open Mission Control to view decisions, outcomes and mission summary.
Coordination Fabric
PartialHow agents register, route and stay in sync.
What: Provides the registry, routing and synchronisation between agents.
Why: Coordination is what turns separate agents into one system.
Next: Open Coordination to see registry and routing internals.
Log Copilot
WorkingAsk questions about your logs in plain English.
What: Answers log questions with cited log lines and a risk level.
Why: Finding the needle in the logs is half of incident response.
Next: Open Logs to ask the copilot a question.
Ghost Development Observatory
PrototypeSee what Ghost is building, thinking with, remembering, and learning — and who owns the next step.
What: Read-only pipeline, release-gate, cognition, memory, and learning snapshot. No merge, shell, or source-write controls.
Why: The founder should not have to read raw JSON to know whether a mission is blocked, stale, or waiting on audit.
Next: Open Development Observatory, then Cognition Lab for detailed memory tests.
Ghost Operator
PrototypeEnter a task. Ghost observes, plans, asks approval, acts, verifies, and learns — supervised.
What: Local observe→plan→approve→act→verify→learn loop. Stop Ghost is absolute. Sensitive targets blocked.
Why: The founder should see what Ghost sees, what it plans, whether it has permission, and whether the action actually worked.
Next: Open Ghost Operator on the dedicated runtime, then run the CIC Notepad marker test.
Ghost Console (Operate)
PrototypeType a goal and watch Ghost run the operating loop end to end.
What: Classifies your goal, builds a short plan, recalls data, scores risk via brain/decide, and stops at approval (read-only preview — execution disabled).
Why: This is the front door that turns Ghost from dashboards into an operational system you can actually drive.
Next: Open Ghost Console and run a goal like “Check the system and tell me what to do next.”
Entity v5.0 Supervised Operator
PrototypeAsk Entity to analyze intent and show a supervised proposal timeline — nothing executes.
What: Authenticated Mission Control panel over Entity v5.0 unified-operator demo: intent → analysis → proposed next action with honesty labels.
Why: Makes existing Entity capability visible and inspectable without Relay auto-dispatch or autonomous execution.
Next: Open Operate and use “Analyze / generate proposal” on the Entity v5.0 panel.
Knowledge Engine
WorkingThe self-documenting knowledge graph behind every capability.
What: Auto-generates documentation, lessons, relationships, search and history for every capability in the system.
Why: Keeps the whole product explainable forever with zero manual documentation work — nothing stays undocumented.
Next: Open the Knowledge Engine to explore the capability graph and architecture.
Demonstrations
5 capabilitiesGuided Demo — Predict Failure
WorkingA scripted walk-through of failure prediction.
What: Steps through how Pulse predicts a failure before it happens.
Why: The fastest way to understand the product in two minutes.
Next: Start the guided failure-prediction demo.
AI Demo — Explain Failure
WorkingWatch the AI explain a failure in plain English.
What: Shows the explanation engine narrating a root cause.
Why: Explanations are what make AI output usable by humans.
Next: Start the explain-failure demo.
Trading Demo
WorkingSee the trading brain make a simulated call.
What: Runs the trading decision flow on live market signals (simulated execution).
Why: Concrete proof of the decision engine in a real domain.
Next: Open Markets and trigger a trading decision.
Mission Demo — Corrective Action
WorkingWatch a mission drive a corrective action.
What: Steps through detecting a problem and taking a corrective action.
Why: Shows the full loop from detection to response.
Next: Start the corrective-action demo.
Talk to Ghost
PrototypeThe product front door — conversation, missions, approvals, and verified results in one place.
What: Authenticated conversational Ghost with Executive cognition, durable session continuity, mission cards, human-readable approvals, completion-truth verification, and honesty labels.
Why: This is the core customer surface of the Ghost Founder Pilot dedicated advisory seat.
Next: Open it and ask Ghost a question, or hand it something to work on.