Evidence before narrative
New Runtime reports verified attribution, probable correlation, and unknown-origin movement separately. A before/after delta is never described as causal without a matching receipt or referral.
Outbound attention · attributable return
Presence
Privacy-safe evidence for how New Runtime leaves its own domain, becomes reusable elsewhere, and earns attributable return attention.
The loop
A page must travel and return evidence
Publication volume is not the result. A useful loop connects a canonical page to an approved artifact and a measurable return.
- 01Evidence change
A source-backed record becomes worth revisiting.
- 02Canonical page
HTML, Markdown, JSON, sources, and a stable URL agree.
- 03Approved packet
Platform copy and visual pass the owner gate.
- 04External surface
Feed, social post, dataset, MCP registry, or verified Webmention.
- 05Return signal
Referral, citation, fork, subscription, download, or agent read.
Before baseline · 2026-07-30T20:30:14.912Z
The gap is attributable human return
These are aggregate observations, not unique people. Scanners, crawlers, and New Runtime monitoring are never promoted into audience numbers.
- Human page views
- 802
- Human search referrals
- 1
- Human AI referrals
- 0
- AI user-agent fetches
- 9
- Machine route requests
- 1,094
- RSS items
- 88
Reusable authority assets
Versioned datasets, feeds, and MCP resources make the work forkable without exposing private discovery infrastructure.
Open the first assetOwner gates stay visible
- Telegram publication requires an OpenClaw private preview, explicit owner approval, and a numeric Telegram message_id.
- X and Bluesky publication require explicit owner approval and a stored provider URL or receipt.
- GitHub Release, Hugging Face, and MCP Registry publication require the matching owner identity and an immutable external receipt.