Git Crafter

4 reposAPI data 2026-11-21 13:36 UTCsite data 2026-11-21 13:36 UTCbuilt 2026-11-21 13:37 UTC

Ideal state

A process is a map of at most six steps where quality, speed and consumption are decided, each scored by one Quality metric.

2 sources, newest 2026-11-04
  1. Claude Code: best practices 2026-11-04
  2. DORA's software delivery performance metrics 2026-01-05

Delivery

A change to Git Crafter, from the plan the owner approves to merged on main

Ideal state

Small changes, planned first, green on first push, reviewed at the head and merged within a day, judged by lead time and failure, not volume.

4 sources, newest 2026-11-04
  1. Claude Code: best practices 2026-11-04
  2. Google engineering practices: speed of code reviews 2026-11-04
  3. Trunk Based Development: short-lived feature branches 2026-11-04
  4. DORA's software delivery performance metrics 2026-01-05
1

Plan

critical path

Quality

100%

PRs linking an issue

goal ≥ 100%

100%
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

The session where the owner approves the plan orchestrates and starts every session it can foresee in that turn.

skills/merging/orchestrating.md#orchestrating
  • Plan mode
  • the owner approves in chat
  • App sessions, each in its own worktree
2

Build

critical path

Quality

0%

Share over 400 lines

goal ≤ 10%

0%
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

Claude builds and reviews; open-weight models do the cheap bulk work.

rules/compute.md
  • pre-push runs no tests; CI runs them (site decision 93)
  • Claude Code
  • Explore agent on Haiku for wide reads
  • OpenRouter free models
  • Ollama cloud
3

Check

critical path

Consumption

0.23 dollars

Runner dollars at list price, the latest merged PR

goal ≤ 0.36 dollars

0.23 dollars
8 days, 2026-11-13 to 2026-11-20

No per-repo data for this step.

CI, test suites and builds run on GitHub Actions; the Mac runs only rare, small, quick things.

rules/compute.md
  • GitHub Actions, standard runners, ubuntu-slim for light jobs
  • Included Actions usage, 300 USD, a budget that stops at it
  • Free lanes: Blacksmith and Ubicloud free minutes, public repos
  • Pinned actions
  • The app's PR monitor with Auto-fix
4

Review

critical path

Quality

40%

Fixed after the first review

goal ≤ 25%

40%
8 days, 2026-11-13 to 2026-11-20

No per-repo data for this step.

A reviewer runs the checks on the head commit and reports no blocking finding; a fix needs a new review.

skills/merging/reviewing.md#reviewing
  • A review pinned to the reviewed head SHA, clean or blocking (site decision 95)
  • Deny and hookify test cases run in claude_deny.py --self-test and hookify_check.py --self-test; shell probes run in CI (rules/gates.md)
  • One pr-review-toolkit reviewer per PR
  • Head SHA pin
  • gh pr checks
5

Merge

critical path

Speed

10.8 hours

Latest PR's hours to merge

goal ≤ 4 hours

10.8 hours
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

The orchestrator merges the reviewed head SHA when every check is green; never --auto or --admin.

skills/merging/SKILL.md#who-merges
  • Merge with the literal reviewed SHA, typed, never read at merge time
  • gh pr merge with --match-head-commit
  • Dependabot bumps each caller's pin

Deploy

A merged change to Git Crafter, from the Pages build to live, and back when it fails

Ideal state

Merges deploy in minutes from a CI build, rarely fail, roll back within the hour, read per service for lead time, failure and recovery.

2 sources, newest 2026-07-15
  1. Cloudflare Workers: rollbacks 2026-07-15
  2. DORA's software delivery performance metrics 2026-01-05
1

Build

critical path

Consumption

10 min

CI minutes behind the latest deploy

goal ≤ 20 min

10 min
8 days, 2026-11-13 to 2026-11-20

No per-repo data for this step.

Every merge to main builds and publishes the pages on its own; nobody deploys by hand.

https://dora.dev/guides/dora-metrics/
  • pages.yml on every push to main
2

Ship

critical path

Speed

1.8 minutes

Pipeline: the latest push to all green

goal ≤ 4 minutes

1.8 minutes
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

A merged change is live once its push goes green, within minutes.

https://dora.dev/guides/dora-metrics/
  • GitHub Pages
3

Verify

critical path

Quality

0%

Failed deploys, last 10

goal ≤ 5%

0%
8 days, 2026-11-13 to 2026-11-20

No per-repo data for this step.

A publish whose checks fail, or pass only on a rerun, counts as a failed change.

https://dora.dev/guides/dora-metrics/
  • pages.yml build checks
  • visual-check screenshots

Board

A feature, from a card on the project board to Done with its merged pull request

Ideal state

One fresh session takes one visual card to one merged pull request; open pull requests carry the limit, and nothing sits in progress unbuilt.

5 sources, newest 2026-11-10
  1. GitHub Docs: Best practices for Projects 2026-11-10
  2. Claude Code Docs: Manage costs effectively 2026-11-10
  3. Anthropic: How we built our multi-agent research system 2026-11-10
  4. DORA: Work in process limits 2026-11-10
  5. DORA: Working in small batches 2025-12-08
1

Card

critical path

Quality

52.4%

Cards read at a glance

goal ≥ 100%

52.4%
8 days, 2026-11-13 to 2026-11-20

No per-repo data for this step.

A feature is one card: a picture of its metric and steps, a table, and a pass-or-fail check.

https://code.claude.com/docs/en/best-practices
  • A draft card on the project board
  • quality/cards/picture.mjs
2

Split

critical path

Speed

0 PRs

Open PRs with a conflict

goal ≤ 0 PRs

0 PRs
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

Cut the work so each card lands on its own and shares no file with another card in progress.

https://www.anthropic.com/research/multi-agent-research-system
  • The Area field
  • Pull requests of 400 lines or fewer
3

Start

critical path

Quality

1 cards

In progress, no open PR

goal ≤ 0 cards

1 cards
8 days, 2026-11-13 to 2026-11-20

No per-repo data for this step.

A fresh session takes one card; the card is in progress only while its pull request is open.

https://code.claude.com/docs/en/costs
  • One app session per card, in its own worktree
  • Workflow: Pull request linked to issue
4

Finish

critical path

Speed

22 PRs

Open PRs now

goal ≤ 10 PRs

18 PRs
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

Finish before starting: merge a green pull request before opening another.

https://dora.dev/capabilities/wip-limits/
  • The app's PR monitor with Auto-fix
  • gh pr merge --match-head-commit
5

Done

critical path

Consumption

–

Runner minutes per pull request, the latest card

goal ≤ 60 minutes

#96, the latest merged pull request on the latest card, opened 2026-11-13 14:47 UTC, before 2026-11-13 16:26 UTC, where the runs read stop covering every workflow

26 minutes
6 days, 2026-11-14 to 2026-11-19

No per-repo data for this step.

The merge moves the card to Done with its pull request linked, within the CI minutes goal; the session ends.

https://docs.github.com/en/issues/planning-and-tracking-with-projects/automating-your-project/using-the-built-in-automations
  • Workflows: Pull request merged, Item closed
  • Auto-archive the session on merge
6

Watch

Speed

7.9 days

Oldest card in progress

goal ≤ 7 days

7 days
8 days, 2026-11-13 to 2026-11-20

No per-repo data for this step.

No card sits in progress past a week; a waiting card returns to Todo and names what it waits on.

https://dora.dev/capabilities/working-in-small-batches/
  • The Stuck view
  • Insights charts

Adopt Git Crafter

A repo, from copying the starters to running Git Crafter's CI and pages and staying current

Ideal state

A repo copies two starters and three CI files, pins Git Crafter to a full SHA, requires its check, and takes each release within days.

4 sources, newest 2026-11-05
  1. GitHub Actions: reuse workflows 2026-11-05
  2. GitHub Actions: secure use reference 2026-11-05
  3. GitHub: about rulesets 2026-11-05
  4. GitHub: Dependabot version updates 2026-11-05
1

Copy

critical path

Quality

100%

Actions pinned to a SHA

goal ≥ 100%

100%
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

Copy the starters, CI and pages callers and Dependabot entry; pin Git Crafter to a full SHA with its version.

https://docs.github.com/en/actions/reference/security/secure-use
  • templates/quality/config.json
  • templates/process/config.json
  • templates/ci/ci.yml
  • templates/ci/pages.yml
  • templates/ci/dependabot.yml

Measure

A repo's numbers, from collected to scored, published, goaled and fixed

Ideal state

Metrics are read live in CI, scored against sourced goals, published on every push, and each miss becomes a labeled issue a pull request closes.

4 sources, newest 2026-11-05
  1. Google SRE book: service level objectives 2026-11-05
  2. DORA's software delivery performance metrics 2026-01-05
  3. GitHub Pages: custom workflows 2026-11-05
  4. GitHub: link a pull request to an issue 2026-11-05
2

Score

critical path

Quality

0 dates

Dated claims older than their area

goal ≤ 0 dates

0 dates
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

A metric meets, nears or misses a goal that names its source, and every figure carries a recent read date.

https://sre.google/sre-book/service-level-objectives/
  • quality/score.mjs
  • Each goal's goal_source
4

Goal

critical path

Quality

74.4%

Issues with an area label

goal ≥ 100%

80%
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

Each miss becomes a typed issue with an area label if one pull request fixes it, else a board card.

https://sre.google/sre-book/service-level-objectives/
  • The work queue
  • Area labels

Code

Git Crafter's own code, from measured to simplified, deleted, explained, checked and held small

Ideal state

Code stays small and plain: the worst function is split first, unused lines are deleted, shared functions say what they do.

6 sources, newest 2026-11-10
  1. Microsoft Learn: Code metrics, cyclomatic complexity 2026-10-20
  2. GitHub Docs: Code Quality metrics and scores reference 2026-11-10
  3. Claude Code Docs: Tools reference 2026-11-10
  4. Anthropic: Effective context engineering for AI agents 2025-09-29
  5. PEP 257: Docstring Conventions 2024-04-17
  6. Claude Code Docs: Best practices for Claude Code 2026-11-10
4

Explain

Quality

100%

Shared functions with a header comment

goal ≥ 100%

100%
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

Every shared function says what it does where it starts.

https://peps.python.org/pep-0257/
  • A header comment or docstring

Triage

A finding in Git Crafter, from intake to the issue or card a merged pull request closes

Ideal state

A finding becomes a typed issue closed within a day, or a card on the board; each closes through its merged pull request.

5 sources, newest 2026-11-10
  1. GitHub Actions billing 2026-11-04
  2. GitHub Docs: About issues 2026-11-10
  3. GitHub Docs: Linking a pull request to an issue 2026-11-10
  4. GitHub Docs: Best practices for Projects 2026-11-10
  5. GitHub Docs: Using the built-in automations 2026-11-10
1

Intake

critical path

Quality

100%

Issue forms in order

goal ≥ 100%

100%
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

Every finding arrives through an issue form with a type, so it is ready to act on.

SUPPORT.md
  • Issue forms
  • Issue types
2

Issue

critical path

Speed

10.8 hours

Latest issue's hours to close

goal ≤ 24 hours

10.8 hours
8 days, 2026-11-13 to 2026-11-20

No per-repo data for this step.

A bug or quick fix one small pull request can close in a day is a typed issue.

CONTRIBUTING.md
  • Typed issue
  • One small pull request
3

Card

critical path

Speed

0.5 hours

Latest card's hours to done

goal ≤ 72 hours

0.5 hours
7 days, 2026-11-14 to 2026-11-20

No per-repo data for this step.

A feature is a card on the board, moved by the board's automations.

CONTRIBUTING.md
  • Project board
  • Built-in automations

Document

A doc, from written visual first to linted, mapped to its files, published, read and loaded

Ideal state

Every fact a reader needs is present, correct and current, in short, visual docs a person finishes in two minutes and Claude loads cheaply.

5 sources, newest 2026-11-10
  1. GitHub Docs: Creating diagrams 2026-11-10
  2. GitHub Docs: Style guide 2026-11-10
  3. GitHub Docs: About the repository README file 2026-11-10
  4. Microsoft Style Guide: Scannable content 2023-06-20
  5. Claude Code Docs: Best practices 2026-11-10
6

Load

Consumption

0 tokens

Tokens past the reading goal

goal ≤ 0 tokens

0 tokens
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

Claude loads no tokens past a doc's reading goal.

https://code.claude.com/docs/en/best-practices
  • Agent docs: CLAUDE.md, skills and rules

Secure

Git Crafter's security, from keys no session reads to alerts closed within a week

Ideal state

No session reads a key, scans run in minutes, the threat model stays current, reports have a private path, and alerts close within a week.

4 sources, newest 2026-11-10
  1. Claude Code: configure permissions 2026-11-04
  2. GitHub Docs: GitHub security features 2026-11-10
  3. GitHub Docs: Quickstart for securing your repository 2026-11-10
  4. GitHub Docs: Adding a security policy to your repository 2026-11-03
1

Keys

critical path

Quality

1 on/off

Secret scanning and push protection

goal ≥ 1 on/off

1 on/off
8 days, 2026-11-13 to 2026-11-20

No per-repo data for this step.

Push protection blocks a commit that holds a key before it reaches the repo.

rules/keys.md
  • Secret scanning
  • Push protection
3

Model

critical path

Quality

1 on/off

Threat model current with its code

goal ≥ 1 on/off

1 on/off
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

The threat model is reviewed whenever the attack surface changes.

THREAT-MODEL.md
  • THREAT-MODEL.md
4

Report

critical path

Quality

1 on/off

Security policy

goal ≥ 1 on/off

1 on/off
9 days, 2026-11-12 to 2026-11-20

No per-repo data for this step.

A security policy gives reporters a private path.

SECURITY.md
  • SECURITY.md
  • Private vulnerability reporting
5

Alerts

critical path

Speed

8.5 days

Oldest open alert

goal ≤ 7 days

7.5 days
8 days, 2026-11-13 to 2026-11-20

No per-repo data for this step.

An open alert closes within a week.

SECURITY.md
  • Dependabot alerts

Ideal state

Every repo runs one CI standard on hosted runners, pinned to a SHA, and skips unchanged work so included minutes last the month.

2 sources, newest 2026-11-04
  1. GitHub Actions: reuse workflows 2026-11-04
  2. GitHub: about billing for GitHub Actions 2026-11-04

The CI standard

Runner dollars at list price, the latest merged PR

0.23 dollars

claude-review0.01 dollars
cloud-test0.01 dollars
pages0.12 dollars
pr-body0.01 dollars
secret-scan0.01 dollars
verify0.07 dollars

templates/ci/README.md

.github6 workflows

The slowest step of each stage95% or more pass85 to 94%Under 85%Not measuredRuns on the laptopA personBlue bar: median minutes

  1. Pull request checks0.2 min

    • dependabot-automerge
      Median
      0.2 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every pull request: automerge on a runner chosen at run time
      Defined in
      .github/workflows/dependabot-automerge.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
    • gitleaks
      Median
      0.2 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every pull request: gitleaks on a runner chosen at run time
      Defined in
      .github/workflows/gitleaks.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
  2. Review

    • Pinned review and merge
      Median
      Not timed
      Pass rate
      not measured
      What runs
      The orchestrator reviews the head SHA, pins a review to it, and merges with gh pr merge --match-head-commit and that literal SHA
      Defined in
      code-os rules/delegation.md
      Kind
      A step a person takes
      Path
      The slowest step in its stage, on the rail
  3. Merge to main0.2 min

    • gitleaks
      Median
      0.2 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every push to main: gitleaks on a runner chosen at run time
      Defined in
      .github/workflows/gitleaks.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
Every step as a table
StageStepKindMedianPass rate
Pull request checksdependabot-automerge (rail)Workflow0.2 min100%
Pull request checksgitleaksWorkflow0.2 min100%
ReviewPinned review and merge (rail)Personnot timednot measured
Merge to maingitleaks (rail)Workflow0.2 min100%

Apart from the pipeline

The slowest step of each stage95% or more pass85 to 94%Under 85%Not measuredRuns on the laptopA personBlue bar: median minutes

  1. Scheduled0.3 min

    • gitleaks-release
      Median
      0.3 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On a schedule: bump on ubuntu-24.04
      Defined in
      .github/workflows/gitleaks-release.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
  2. On demand0.2 min

    • gitleaks-release
      Median
      0.2 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On dispatch: bump on ubuntu-24.04
      Defined in
      .github/workflows/gitleaks-release.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
  3. Called by other repos

    • package-ci
      Median
      Not timed
      Pass rate
      not measured
      What runs
      On a call from another workflow: route on soulcraft-router and gitleaks calls shared gitleaks.yml, then verify on a runner the Blacksmith guard picks
      Defined in
      .github/workflows/package-ci.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
    • package-gpu
      Median
      Not timed
      Pass rate
      not measured
      What runs
      On a call from another workflow: gpu on self-hosted, gpu-pod
      Defined in
      .github/workflows/package-gpu.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • package-release
      Median
      Not timed
      Pass rate
      not measured
      What runs
      On a call from another workflow: tag on a runner chosen at run time, then gate calls shared package-ci.yml and guard on a runner chosen at run time, then release on a runner chosen at run time
      Defined in
      .github/workflows/package-release.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
Every step as a table
StageStepKindMedianPass rate
Scheduledgitleaks-release (rail)Workflow0.3 min100%
On demandgitleaks-release (rail)Workflow0.2 min100%
Called by other repospackage-ci (rail)Workflownot timednot measured
Called by other repospackage-gpuWorkflownot timednot measured
Called by other repospackage-releaseWorkflownot timednot measured

gh api graphql: HEAD:.github/workflows, HEAD:.githooks and HEAD:.github/cloud-test.json; each workflow's latest run timed, its pass rate over its last 20 runs read 2026-11-21 13:37 UTC

comfyui-base6 workflows

The slowest step of each stage95% or more pass85 to 94%Under 85%Not measuredRuns on the laptopA personBlue bar: median minutes

  1. Pull request checks15.6 min

    • sage-wheels
      Median
      15.6 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every pull request that changes its inputs: resolve on ubuntu-24.04, then build on ubuntu-24.04, then publish on ubuntu-24.04
      Defined in
      .github/workflows/sage-wheels.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
    • dependabot-automerge
      Median
      0.2 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every pull request: automerge calls shared dependabot-automerge.yml
      Defined in
      .github/workflows/dependabot-automerge.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • download-throughput
      Median
      0.4 minutes
      Pass rate
      92.3% of runs passed; cancelled runs are left out
      What runs
      On every pull request that changes its inputs: throughput on ubuntu-24.04
      Defined in
      .github/workflows/download-throughput.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • gitleaks
      Median
      0.3 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every pull request: gitleaks calls shared gitleaks.yml
      Defined in
      .github/workflows/gitleaks.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • verify
      Median
      6.1 minutes
      Pass rate
      75% of runs passed; cancelled runs are left out
      What runs
      On every pull request: verify on ubuntu-24.04
      Defined in
      .github/workflows/verify.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
  2. Review

    • Pinned review and merge
      Median
      Not timed
      Pass rate
      not measured
      What runs
      The orchestrator reviews the head SHA, pins a review to it, and merges with gh pr merge --match-head-commit and that literal SHA
      Defined in
      code-os rules/delegation.md
      Kind
      A step a person takes
      Path
      The slowest step in its stage, on the rail
  3. Merge to main6.9 min

    • verify
      Median
      6.9 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every push to main: verify on ubuntu-24.04
      Defined in
      .github/workflows/verify.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
    • gitleaks
      Median
      0.2 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every push to main: gitleaks calls shared gitleaks.yml
      Defined in
      .github/workflows/gitleaks.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • release
      Median
      0.2 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every push to main: release on ubuntu-24.04
      Defined in
      .github/workflows/release.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
Every step as a table
StageStepKindMedianPass rate
Pull request checkssage-wheels (rail)Workflow15.6 min100%
Pull request checksdependabot-automergeWorkflow0.2 min100%
Pull request checksdownload-throughputWorkflow0.4 min92.3%
Pull request checksgitleaksWorkflow0.3 min100%
Pull request checksverifyWorkflow6.1 min75%
ReviewPinned review and merge (rail)Personnot timednot measured
Merge to mainverify (rail)Workflow6.9 min100%
Merge to maingitleaksWorkflow0.2 min100%
Merge to mainreleaseWorkflow0.2 min100%

gh api graphql: HEAD:.github/workflows, HEAD:.githooks and HEAD:.github/cloud-test.json; each workflow's latest run timed, its pass rate over its last 20 runs read 2026-11-21 13:37 UTC

compute-crafter2 workflows

The slowest step of each stage95% or more pass85 to 94%Under 85%Not measuredRuns on the laptopA personBlue bar: median minutes

  1. Pull request checks0.9 min

    • ci
      Median
      0.9 minutes
      Pass rate
      85.7% of runs passed; cancelled runs are left out
      What runs
      On every pull request: ci calls SoulCraft-Org/git-crafter/.github/workflows/ci-standard.yml
      Defined in
      .github/workflows/ci.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
    • pages
      Median
      0.4 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every pull request: build on ubuntu-latest
      Defined in
      .github/workflows/pages.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
  2. Review

    • Pinned review and merge
      Median
      Not timed
      Pass rate
      not measured
      What runs
      The orchestrator reviews the head SHA, pins a review to it, and merges with gh pr merge --match-head-commit and that literal SHA
      Defined in
      code-os rules/delegation.md
      Kind
      A step a person takes
      Path
      The slowest step in its stage, on the rail
  3. Merge to main0.9 min

    • ci
      Median
      0.9 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every push to main: ci calls SoulCraft-Org/git-crafter/.github/workflows/ci-standard.yml
      Defined in
      .github/workflows/ci.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
    • pages
      Median
      0.7 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every push to main: build on ubuntu-latest, then deploy on ubuntu-slim
      Defined in
      .github/workflows/pages.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
Every step as a table
StageStepKindMedianPass rate
Pull request checksci (rail)Workflow0.9 min85.7%
Pull request checkspagesWorkflow0.4 min100%
ReviewPinned review and merge (rail)Personnot timednot measured
Merge to mainci (rail)Workflow0.9 min100%
Merge to mainpagesWorkflow0.7 min100%

gh api graphql: HEAD:.github/workflows, HEAD:.githooks and HEAD:.github/cloud-test.json; each workflow's latest run timed, its pass rate over its last 20 runs read 2026-11-21 13:37 UTC

git-crafter12 workflows

The slowest step of each stage95% or more pass85 to 94%Under 85%Not measuredRuns on the laptopA personBlue bar: median minutes

  1. Pull request checks2 min

    • pages
      Median
      2 minutes
      Pass rate
      70% of runs passed; cancelled runs are left out
      What runs
      On every pull request that changes its inputs: runner on ubuntu-slim and changes on ubuntu-slim, then docs-gate on ubuntu-slim and code-gate on ubuntu-slim and action-render on a runner chosen at run time
      Defined in
      .github/workflows/pages.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
    • branch-cleanup
      Median
      Not timed
      Pass rate
      not measured
      What runs
      On every pull request: branch-cleanup on ubuntu-slim
      Defined in
      .github/workflows/branch-cleanup.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • claude-review
      Median
      0.2 minutes
      Pass rate
      90% of runs passed; cancelled runs are left out
      What runs
      On every pull request that changes its inputs: gate on ubuntu-slim, then review on ubuntu-latest
      Defined in
      .github/workflows/claude-review.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • cloud-test
      Median
      1.1 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every pull request that changes its inputs: cloud-test on ubuntu-slim
      Defined in
      .github/workflows/cloud-test.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • pr-body
      Median
      0.3 minutes
      Pass rate
      85% of runs passed; cancelled runs are left out
      What runs
      On every pull request: pr-body on ubuntu-slim
      Defined in
      .github/workflows/pr-body.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • secret-scan
      Median
      0.3 minutes
      Pass rate
      90% of runs passed; cancelled runs are left out
      What runs
      On every pull request: secret-scan on ubuntu-slim
      Defined in
      .github/workflows/secret-scan.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • verify
      Median
      1.8 minutes
      Pass rate
      90% of runs passed; cancelled runs are left out
      What runs
      On every pull request: runner calls ./.github/workflows/ci-overflow.yml and dependency-review on ubuntu-slim, then ci calls ./.github/workflows/ci-standard.yml
      Defined in
      .github/workflows/verify.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
  2. Review

    • Pinned review and merge
      Median
      Not timed
      Pass rate
      not measured
      What runs
      The orchestrator reviews the head SHA, pins a review to it, and merges with gh pr merge --match-head-commit and that literal SHA
      Defined in
      code-os rules/delegation.md
      Kind
      A step a person takes
      Path
      The slowest step in its stage, on the rail
  3. Merge to main3.1 min

    • pages
      Median
      3.1 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every push to main: runner on ubuntu-slim and docs-gate on ubuntu-slim and code-gate on ubuntu-slim, then action-render on a runner chosen at run time and build on a runner chosen at run time, then deploy on ubuntu-slim
      Defined in
      .github/workflows/pages.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
    • secret-scan
      Median
      0.3 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every push to main: secret-scan on ubuntu-slim
      Defined in
      .github/workflows/secret-scan.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • verify
      Median
      1.4 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every push to main: runner calls ./.github/workflows/ci-overflow.yml, then ci calls ./.github/workflows/ci-standard.yml
      Defined in
      .github/workflows/verify.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
    • wiki
      Median
      0.2 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On every push to main that changes its inputs: wiki on ubuntu-slim
      Defined in
      .github/workflows/wiki.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
Every step as a table
StageStepKindMedianPass rate
Pull request checkspages (rail)Workflow2 min70%
Pull request checksbranch-cleanupWorkflownot timednot measured
Pull request checksclaude-reviewWorkflow0.2 min90%
Pull request checkscloud-testWorkflow1.1 min100%
Pull request checkspr-bodyWorkflow0.3 min85%
Pull request checkssecret-scanWorkflow0.3 min90%
Pull request checksverifyWorkflow1.8 min90%
ReviewPinned review and merge (rail)Personnot timednot measured
Merge to mainpages (rail)Workflow3.1 min100%
Merge to mainsecret-scanWorkflow0.3 min100%
Merge to mainverifyWorkflow1.4 min100%
Merge to mainwikiWorkflow0.2 min100%

Apart from the pipeline

The slowest step of each stage95% or more pass85 to 94%Under 85%Not measuredRuns on the laptopA personBlue bar: median minutes

  1. Called by other repos

    • ci-overflow
      Median
      Not timed
      Pass rate
      not measured
      What runs
      On a call from another workflow: runner on ubuntu-slim
      Defined in
      .github/workflows/ci-overflow.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
    • ci-standard
      Median
      Not timed
      Pass rate
      not measured
      What runs
      On a call from another workflow: changes on ubuntu-slim, then short on ubuntu-slim and standard on a runner the Blacksmith guard picks, then ci-ok on ubuntu-slim
      Defined in
      .github/workflows/ci-standard.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
  2. On demand0.8 min

    • scorecard
      Median
      0.8 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On dispatch: analysis on ubuntu-latest
      Defined in
      .github/workflows/scorecard.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
    • push-ruleset-proof
      Median
      Not timed
      Pass rate
      not measured
      What runs
      On dispatch: probe on ubuntu-slim
      Defined in
      .github/workflows/push-ruleset-proof.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
  3. Scheduled0.6 min

    • scorecard
      Median
      0.6 minutes
      Pass rate
      100% of runs passed; cancelled runs are left out
      What runs
      On a schedule: analysis on ubuntu-latest
      Defined in
      .github/workflows/scorecard.yml
      Kind
      An automated workflow: a GitHub Actions workflow or a Cloudflare build
      Path
      The slowest step in its stage, on the rail
Every step as a table
StageStepKindMedianPass rate
Called by other reposci-overflow (rail)Workflownot timednot measured
Called by other reposci-standardWorkflownot timednot measured
On demandscorecard (rail)Workflow0.8 min100%
On demandpush-ruleset-proofWorkflownot timednot measured
Scheduledscorecard (rail)Workflow0.6 min100%

gh api graphql: HEAD:.github/workflows, HEAD:.githooks and HEAD:.github/cloud-test.json; each workflow's latest run timed, its pass rate over its last 20 runs read 2026-11-21 13:37 UTC

Ideal state

Many fast unit tests, fewer broad ones, few end-to-end; every suite runs in CI on every PR, and main stays green.

1 source, newest 2018-02-26
  1. Martin Fowler: The Practical Test Pyramid 2018-02-26

No suite runs on a local machine. A session dispatches one to cloud-test.

.github2 suites
SuiteWhenChecksMedianPass
package-cia call from another workflowroom on the runner (drop the image's unused toolchains); what this pull request changes (notes only skip the gate); ComfyUI Base; Chromium for the App-view tier (each playwright version the package's _build pins); 7 morenot timednot measured
verifya call from another workflowroom on the runner (drop the image's unused toolchains); what this pull request changes (notes only skip the gate); ComfyUI Base; Chromium for the App-view tier (each playwright version the package's _build pins); 7 more––
package-releasea call from another workflow; skips its work when nothing it checks changedthe cast guard is required for a release; the meta repo (org variable META_REPO, owner/name; the app token wants the bare name); soulcraft-ci token (reads the private meta repo for the cast guard); the meta repo's cast tool; 1 morenot timednot measured
guarda call from another workflowthe cast guard is required for a release; the meta repo (org variable META_REPO, owner/name; the app token wants the bare name); soulcraft-ci token (reads the private meta repo for the cast guard); the meta repo's cast tool; 1 more––

gh api graphql: HEAD:.github/workflows, HEAD:.githooks and HEAD:.github/cloud-test.json; each workflow's latest run timed, its pass rate over its last 20 runs read 2026-11-21 13:37 UTC

comfyui-base1 suite
SuiteWhenChecksMedianPass
verifyevery push to main; every pull request; dispatchverify6.1 min75%

gh api graphql: HEAD:.github/workflows, HEAD:.githooks and HEAD:.github/cloud-test.json; each workflow's latest run timed, its pass rate over its last 20 runs read 2026-11-21 13:37 UTC

compute-crafter1 suite
SuiteWhenChecksMedianPass
cievery pull request; every push to mainci calls SoulCraft-Org/git-crafter/.github/workflows/ci-standard.yml0.9 min85.7%

gh api graphql: HEAD:.github/workflows, HEAD:.githooks and HEAD:.github/cloud-test.json; each workflow's latest run timed, its pass rate over its last 20 runs read 2026-11-21 13:37 UTC

git-crafter2 suites
SuiteWhenChecksMedianPass
secret-scanevery pull request; every push to main; dispatchSelf-test (the scan's range, with a planted fake key); Scan the commits this run answers for0.3 min90%
verifyevery pull request; every push to mainrunner calls ./.github/workflows/ci-overflow.yml and dependency-review on ubuntu-slim, then ci calls ./.github/workflows/ci-standard.yml1.8 min90%

cloud-test verify self-test:deny-rules self-test:hookify-rules self-test:voice-lint self-test:sync-tool self-test:comfy-guard

gh api graphql: HEAD:.github/workflows, HEAD:.githooks and HEAD:.github/cloud-test.json; each workflow's latest run timed, its pass rate over its last 20 runs read 2026-11-21 13:37 UTC