Skip to content

Trial page: empty and repeated parts crowd out what a trial recorded, and a record cannot be saved or copied whole #179

Description

@aitianshi

Priority

high. Mostly the trial page's templates and styles; it reaches the backend in app/nullius/runs/artifacts.py (a record whole) and runs/cli.py (two decimals).

Components

  • app/nullius (backend)
  • app/web (routes)
  • templates and static files
  • docs and the site
  • catalogue (benchmarks, agents, capabilities)
  • the standard (ports, kits, relay, runs)
  • CI, workflows and skills

Today

A trial's page spends its first screen on parts that say nothing, or say it twice:

  • a strip of chips, "Trajectory · recorded", "Trace · recorded", "Reward · recorded", "Rate limits · none logged", which the tabs and the Reward card already say;
  • a "Shape of the run" title over one panel, and under the time bar a paragraph explaining it;
  • cards and panels drawn empty: "Setup —", "Reward —", an empty usage panel, a judge card reading "No record of its call for this one" (nothing in the application writes the record it reads);
  • the tab's own name again as a heading inside the Artifacts, Conversation and judge tabs.

And it hides some of what the trial recorded:

  • durations to one decimal: 0.05 s reads "0.1s", 0.0008 s reads "0.0s";
  • the agent's name is nowhere in the head, only its version;
  • a record on the Artifacts tab can be read one page of 8,000 characters at a time, and neither saved nor copied whole; its Back link returns to Overview.
The trial page today: cards, strip, title, bar and paragraph

Wanted

A page that reads as a summary of the trial: what ran, how it went, where the time went, with room for more tiles. Nothing is drawn for what the trial has not, and nothing is said twice.

Acceptance

  • The evidence strip and the "Shape of the run" title are gone from the trial page.
  • Under the trial's name, one row of blocks says what ran (the benchmark and its runner, or the importer of an imported trial; the agent with its mark, name, version and model) and how it went (the dataset's checks, the reward or the score, the agent's time, the steps). A value the trial has not has no cell and no dash.
  • "Where the time went" is a tile with its total in the heading and no paragraph: a chip for every phase the trial recorded, however short, none for one it did not, and each chip's hover says what the phase holds on that kind of run.
  • A panel with nothing to show is not drawn; the benchmark judge card and the code that read verifier/judge.json are removed.
  • A duration reads to two decimals wherever one is drawn: the pages, the run page's script, a span's minutes, nullius runs show.
  • The Artifacts tab lists only the records the trial wrote, each with Read and Download. A record's page has Copy and Download of the whole record, redacted as the reader shows it, and a back button to the Artifacts tab.
  • The Artifacts, Conversation and judge tabs do not repeat their tab's name as a heading; the judge tabs keep the experimental badge, and say a refusal or a truncated state only when there is one.
  • No sideways scroll and no text running into its neighbour at 390, 820, 1024 and 1280 px, dark and light.

Dependencies and related

The standard's agent setup phase (#180, open; pull request #174) is one of the phases this page draws. This is pull request #173. #173 merges first, #174 after it.

Activity

  1. added
    enhancementNew feature or request
    P3When convenient: wording, layout, tests
    on Nov 17, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    P3When convenient: wording, layout, testsenhancementNew feature or request

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions