Signals — governed relevance for the outside world
HUMΛN Signals is your intelligence layer for the outside world — structured, governed, explainable. It is not a notification center: it is relevance-in-the-loop, tied to HumanOS routing and the same receipts as every other governed output.
Watch what matters → detect change → judge relevance → route to the right workflow or person → generate governed artifacts → learn from what you do with them.
The seven-step intelligence loop
- Source — Where HUMΛN watches — sites, APIs, feeds, repos, uploads.
- Observation — What changed — normalized events, not noisy alerts.
- Signal — What it might mean — structured, typed, confidence-scored.
- Judge — Does it matter for you? For whom? How urgently?
- Route — To Companion, a workflow, review, or escalation — with policy.
- Artifact — Governed outputs — briefs, drafts, battlecards — auditable and attributable.
- Feedback — What you do with it teaches the loop — without silent drift.
Where Signals lives
- Companion — a signal can surface in conversation; you act or dismiss with explicit intent.
- Control Plane — Pulse for triage, Deep Dives for artifacts, watchlists for coverage.
- Workforce Cloud — when a signal needs human review before action, it lands in governed queues.
- Marketplace — reference bundles show how Signals plugs into governed workflows end-to-end.
Not a shadow system
Every Signals artifact carries the same provenance, governance, and challenge rights as other HUMΛN outputs. Intelligence that arrives at a conclusion must be contestable — the same challenge rights that apply to decisions apply here. Signals also demonstrates HumanOS orchestration in production as a reference path — not a sidecar app bolted onto the stack.
Docs & community