Public demo shell

Forgejo Code Server Shell

This is the fastest way to understand the product: Forgejo keeps the code-hosting surface recognizable, while Vaglio adds analysis, provenance, and investor-readable repository narratives beside it.

Start with this demo Open public source Open imported Forgejo target Open snapshot reports
Selected demo details
kubernetes/kubernetes

Large-scale upstream project with many subsystems, strong reviewer specialization, and obvious hotspot surfaces.

  • Shows the Forgejo shell and Vaglio boundary immediately.
  • Highlights a curated dashboard instead of starting with infrastructure controls.
  • Gives an outsider an obvious first click without verbal narration.

Curated Investor Demos

NixOS/nixpkgs GitHub source Massive infra codebase with broad subsystem ownership and frequent coordination churn. kubernetes/kubernetes GitHub source Large-scale upstream project with many subsystems, strong reviewer specialization, and obvious hotspot surfaces. Recommended first click forgejo/forgejo Codeberg source A self-hosted forge codebase that makes the product boundary itself easy to demonstrate.

End-to-End Import Flow

Public source
kubernetes/kubernetes
Imported Forgejo target
vaglio-demos/kubernetes
Import public source done

Mirror kubernetes/kubernetes into a Forgejo-hosted demo namespace without bespoke setup.

Translate Git edge to jj semantics done

Project refs, pull requests, and merges through the Git↔jj gateway before analysis.

Render investor dashboard done

Expose maintainer concentration, expertise signals, hotspots, provenance overlays, and the first stress/heat surface on the same host.

Forgejo Edge

navigation Forgejo repository https://codeberg.org/vaglio-demos/kubernetes navigation Forgejo pull requests https://codeberg.org/vaglio-demos/kubernetes/pulls navigation Vaglio roundtable dashboard / navigation Vaglio analysis surface /forgejo-shell#analysis-surface

Reuse vs Replace Boundary

Repository browsing Forgejo

Use Forgejo's repo, tree, issue, and pull-request UI unchanged.

User and session management Forgejo

Reuse Forgejo authentication and session handling instead of building a parallel account system.

Web chrome Forgejo

Treat Vaglio as an adjacent panel or sidecar surface, not a hard fork of Forgejo chrome.

Webhook ingress Forgejo

Let Forgejo emit Git-shaped events while Vaglio translates them at the gateway boundary.

Analysis and deliberation Vaglio

Keep the internal source of truth `jj`-first behind the Git↔jj translation layer.

Vaglio Analysis Surface

Branch projection

%{
  git: %{kind: :branch_ref, branch: "master", ref: "refs/heads/master"},
  jj: %{
    kind: :bookmark_projection,
    writable: true,
    bookmark: "git/heads/master"
  },
  provenance: %{lossy: false, notes: [], translated_from: :git}
}

Commit projection

%{
  git: %{kind: :commit, sha: "cafebabe"},
  jj: %{kind: :revset_selector, revset: "commit_id(\"cafebabe\")"},
  provenance: %{
    lossy: false,
    notes: ["Git commit SHAs map to explicit `jj` revset selectors, not synthetic change IDs."],
    translated_from: :git
  }
}

Review projection

%{
  git: %{
    state: :open,
    title: "Controller observability and ownership tracing",
    number: 884,
    kind: :pull_request,
    head_ref: "feature/controller-observability",
    base_ref: "master"
  },
  jj: %{
    change_key: "git/pr/884",
    kind: :change_proposal,
    base_bookmark: "git/heads/master",
    head_bookmark: "git/pr/884/head",
    review_state: :open,
    source_branch_bookmark: "git/heads/feature/controller-observability"
  },
  provenance: %{
    lossy: false,
    notes: ["Pull requests are projected as change proposals, not canonical internal review objects."],
    translated_from: :git
  }
}

Merge projection

%{
  git: %{
    strategy: :merge,
    kind: :merge_event,
    base_ref: "master",
    merged_commit_sha: "cafebabe",
    pull_request_number: 884
  },
  jj: %{
    kind: :bookmark_move,
    merge_strategy: :merge,
    bookmark: "git/heads/master",
    target_revset: "commit_id(\"cafebabe\")"
  },
  provenance: %{
    lossy: false,
    notes: ["Non-squash merges preserve Git edge ancestry most directly."],
    translated_from: :git
  }
}

Investor Dashboard

Subsystem sprawl with highly specialized reviewer clusters.

The investor story is not just repo analytics. Vaglio shows where controller, API machinery, and release engineering knowledge is concentrated, and where agent output needs durable provenance to stay governable.

Maintainer concentration
0.47

Top maintainers are less dominant overall, but critical subsystems still hinge on a narrow review core.

Contributor expertise signals
21

A deep bench exists, but expertise is partitioned by subsystem.

Subsystem hotspots
4

API machinery, controllers, test-infra, and release tooling dominate coordination heat.

Expertise signals

SIG-aligned expertise
Contributor clusters map cleanly to subsystem-specific knowledge pockets.
Reviewer continuity
A small set of people repeatedly stabilizes the same risky surfaces across release cycles.

Subsystem hotspots

API machinery — Architectural blast radius
Schema and compatibility changes ripple across client, server, and docs surfaces.
Controllers — High churn density
Frequent logic changes make ownership continuity especially valuable.
Release tooling — Human bottleneck
Investor-facing story: Vaglio shortens the path from change to trustworthy rollout narrative.

Provenance overlays

Roundtable-backed triage
Attach decision history to high-blast-radius changes before they reach release coordination.
Lossy Git merge surfacing
Show where squash/rebase hides lineage so reviewers know when extra explanation is required.

Stress & Change Heat

Active-inference surface
Stress concentrates where subsystem specialization meets release obligations.

The active-inference view highlights controller and API machinery paths where small contributor pools must absorb high ambiguity under strong compatibility constraints.

Derived from sampled branch history
Sampled commits
6719

Shallow history sample from the tracked branch.

Contributor count
562

Unique contributors observed in the sampled window.

Top author share
45.0%

Share of sampled commits attributable to the top three contributors.

Commit cadence
9.7/day

Recent sampled commits per day across the fetched window.

Branch stress
0.69

Cross-SIG changes amplify review contention even when overall ownership is broad.

History heat
4 peaks

API machinery, controllers, release tooling, and test-infra recur as pressure zones.

Active-inference confidence
high

Maintainer specialization patterns are strong enough to make the hotspot ranking legible.

Project Mind Heatmap

Modules with low prediction error read as settled. Repeated coordination pressure, concentrated authorship, and hot paths rise into contested zones.

Settled
Watch
Contested
api machinery path
Contested
Curated subsystem Stress high Heat 0.84
Compatibility obligations create repeated prediction error whenever controller-facing schema assumptions move.
Appraisal value: Investigate and adversarially review
controller logic path
Contested
Curated subsystem Stress high Heat 0.79
Fast-moving control loops create dense change history, but only some of that churn resolves underlying uncertainty.
Appraisal value: Investigate and adversarially review
release tooling path
Watch
Curated subsystem Stress medium Heat 0.66
Stress rises when release readiness depends on a narrow operational review core.
Appraisal value: Track and gather more evidence
pkg/scheduler/schedule_one_podgroup_test.go
Contested
Sampled branch path Stress high Heat 1.0
2 mentions in the sampled branch window. This is where recent change energy and maintainer attention are likely to collide first.
Appraisal value: Investigate and adversarially review
staging/src/k8s.io/apiserver/pkg/cel/common/valuesschemalesstyped_test.go
Contested
Sampled branch path Stress high Heat 1.0
2 mentions in the sampled branch window. This is where recent change energy and maintainer attention are likely to collide first.
Appraisal value: Investigate and adversarially review
pkg/scheduler/framework/plugins/defaultpreemption/default_preemption.go
Contested
Sampled branch path Stress high Heat 1.0
2 mentions in the sampled branch window. This is where recent change energy and maintainer attention are likely to collide first.
Appraisal value: Investigate and adversarially review
api machinery path high stress
Heat score 0.84

Compatibility obligations create repeated prediction error whenever controller-facing schema assumptions move.

controller logic path high stress
Heat score 0.79

Fast-moving control loops create dense change history, but only some of that churn resolves underlying uncertainty.

release tooling path medium stress
Heat score 0.66

Stress rises when release readiness depends on a narrow operational review core.

sampled branch concentration medium stress
Heat score 0.45

The sampled branch history shows 562 contributors with 45.0% of commits concentrated in the top three authors, indicating medium coordination concentration.

History heat timeline

T-30d
Subsystem drift
Independent changes appear safe locally but begin to accumulate cross-cutting review cost.
T-10d
Compatibility crunch
Schema, controller, and release concerns converge on the same merge window.
T-2d
Review compression
A small reviewer set absorbs most of the remaining uncertainty before release.
sampled
9.7/day cadence
Derived from 6719 sampled commits on the tracked branch, using shallow history rather than curated-only narrative.

Sampled Repo Evidence

Tracked branch sample

This section is computed from a shallow sample of the public source branch, not hand-authored demo prose. It is meant to be the first visible proof that the page is tied to actual repository state.

Top sampled contributors

Author Email Commits
Kubernetes Prow Robot 20407524+k8s-ci-robot@users.noreply.github.com 2582
Patrick Ohly patrick.ohly@intel.com 290
Davanum Srinivas davanum@gmail.com 123
yongruilin yongrlin@outlook.com 108
Lalit Chauhan lalitchauhan@google.com 105

Recent sampled commits

When Author SHA
2026-05-26 Kubernetes Prow Robot 48ba4695
2026-05-26 Kubernetes Prow Robot 083b28d8
2026-05-26 Kubernetes Prow Robot b0947365
2026-05-26 Kubernetes Prow Robot c735e8c5
2026-05-26 Kubernetes Prow Robot ec62fa4d
2026-05-26 Bartosz 27c939be
2026-05-26 Maciej Wyrzuc 29ddc290
2026-05-26 Kubernetes Prow Robot 556b2f9d
2026-05-26 Lucas Käldström 27e0fa95
2026-05-25 Kubernetes Prow Robot b859f5a5
2026-05-25 Kubernetes Prow Robot 704d4840
2026-05-25 Kubernetes Prow Robot c3c78e43

Sampled path hotspots

Path Mentions
pkg/scheduler/schedule_one_podgroup_test.go 2
staging/src/k8s.io/apiserver/pkg/cel/common/valuesschemalesstyped_test.go 2
pkg/scheduler/framework/plugins/defaultpreemption/default_preemption.go 2
pkg/scheduler/schedule_one_podgroup.go 2
staging/src/k8s.io/apiserver/pkg/cel/common/valuesschemalesstyped.go 2
staging/src/k8s.io/csi-translation-lib/go.sum 1
staging/src/k8s.io/streaming/go.sum 1
staging/src/k8s.io/client-go/go.mod 1

JJ vs Git Infrastructure Benchmark

Kubernetes subsystem coordination benchmark

For very large projects, Git compatibility should be a border policy, not the internal execution model.

Concurrent changes
48
Ephemeral workspaces
30
Conflict recovery cases
10
Ingest window
3.2 minutes
Reproducible workload hooks
  • SIG ownership trace
  • review lineage overlay
  • release-risk explanation
jj-native core native
Throughput 1.00x Handles many simultaneous controller/API edits more cleanly.
Ingest latency 0.76x Less reconstitution work when syncing agent-produced changes into analysis.
Operational complexity 0.62x Fewer branch-management artifacts to clean up at scale.
Ecosystem compatibility 0.61x Still needs a Git-shaped facade for upstream contributor expectations.
Provenance fidelity 0.96x Preserves high-blast-radius reasoning with less lossy merge translation.
  • Stronger long-term substrate for agent-scale code velocity.
  • Needs deliberate compatibility surfaces for human contributors.
Git-compatible edge-first flow compatible
Throughput 0.74x Review and branch overhead compounds as subsystem concurrency rises.
Ingest latency 1.00x Baseline Git mirror/import flow remains serviceable for demos and adoption.
Operational complexity 1.00x Baseline includes more policy and tooling glue to regain semantics.
Ecosystem compatibility 1.00x Matches what current contributors and integrations already expect.
Provenance fidelity 0.64x More lineage loss under squash/rebase-heavy flows unless Vaglio repairs it.
  • Easier external adoption and procurement story.
  • Less compelling substrate once agents dominate throughput.

Keep native in jj

agent batch execution
Best handled inside the native Vaglio core.
semantic lineage
Best handled inside the native Vaglio core.
cross-subsystem recovery
Best handled inside the native Vaglio core.
benchmark-driven planning
Best handled inside the native Vaglio core.

Keep compatible at the edge

upstream mirrors
Expose through the Forgejo/Git-facing shell for adoption.
pull-request presentation
Expose through the Forgejo/Git-facing shell for adoption.
review notifications
Expose through the Forgejo/Git-facing shell for adoption.
external ecosystem integrations
Expose through the Forgejo/Git-facing shell for adoption.

Extension Seams

Git edge Roundtable.Adapters.Forgejo

Reads and writes discussion repos against Forgejo's API.

Semantic gateway Roundtable.Translation.GitToJj

Maps Git refs, pull requests, commits, and merges into `jj`-native analysis objects.

Vaglio analysis UI RoundtableWeb.DiscussionLive

Hosts the existing roundtable dashboard without requiring Git to become the internal source of truth.

Advanced Prototype Source Controls

These controls are here for operators and demo preparation. Most visitors should start with the recommended curated demo above instead of changing the shell inputs first.