The specification behind TESSERA: a live 2D/3D data matrix for infrastructure you're already running - Docker containers, systemd units, Prometheus-compatible metrics, custom scripts, or the results of an existing Grafana query. One tile per monitored entity, scored 0-100 by how much attention it currently warrants, rendered as a flat grid or an orbitable 3D skyline.
This repository defines the specification: the data model, the scoring
model, and the visual design language every TESSERA implementation
conforms to. The reference implementations - a self-hostable server
(tessera) and a Grafana panel plugin (tessera-grafana-panel) - are
built and published separately; see Repository map.
TESSERA is developed the way a design language is meant to be: one specification, multiple independent surfaces implementing it. Before publishing the reference implementations, we are running a formal visual and structural consistency review across every existing rendering surface, resolving them against a single specification here. Publishing that specification - and the review itself - as an open process, rather than only publishing a finished result later, produces a better outcome: outside review of a design specification catches things an internal review misses, and it lets contributors shape the data model and the visual language before implementations are locked in around it, not after.
This is standard practice for a specification meant to outlast any one
codebase - the same reason formats and protocols publish RFC-style
drafts for comment before implementations ship. vidimus-spec (our
document-format project) follows the same model.
- Scope and status - what TESSERA is and deliberately is not (a companion to Grafana/Prometheus/Uptime Kuma, not a replacement; a read-only viewer, never a control plane; a flat concern matrix, not a topology graph).
- Data model and tile scoring - the tile schema, the shared score-to-color model, and the config-first adapter model that lets a new data source be added without a code change.
- Visual language - the specification every implementation converges on: colors, spacing, motion, and typography, with the values already confirmed identical across implementations, and the values currently under review, both stated plainly.
- Design review agenda - the specific, substantive questions this review is working through, open for comment.
Contributions are welcome on the specification and on the implementations:
- Specification and design discussion happens here - propose a change to the data model, comment on a visual-language value, weigh in on an open design-review item.
- Implementation contributions land in
tesseraandtessera-grafana-panelonce those repositories are public (see Repository map below for current status). - The open design-review agenda is real engineering work, not starter tasks - each item is a specific, substantive question about how TESSERA should render across surfaces, and resolving one is a meaningful contribution to the specification itself.
See CONTRIBUTING.md for the mechanics, the Code of Conduct that applies to all of it, and the AI Policy before using an AI tool to help with a contribution. Every issue and pull request is read directly by the team; changes worth adopting are merged deliberately, never automatically.
| Repository | Purpose | Status |
|---|---|---|
| tessera-spec (this one) | Specification: data model, scoring, visual design language | Published |
| tessera | Reference implementation: self-hostable server and viewer | In development, publishing next |
| tessera-grafana-panel | Reference implementation: Grafana panel plugin | In development, publishing next |
The specification text in this repository is licensed under
CC BY 4.0 (Creative Commons Attribution 4.0 International) -
anyone can read, implement against, and redistribute it, including
commercially, as long as Favus is credited. The reference
implementations (tessera, tessera-grafana-panel) are licensed
separately under AGPLv3 - see their own repositories.
Built by Favus.