vllm-project/semantic-router GitHub Actions scorecard

Public GitHub Actions data, last 30 days. Updated 7/11/2026, 5:18:54 AM.

Data sourced from public GitHub. GitSpider is not affiliated with or endorsed by this repository's owners. Request removal.

40 min/mo
recoverable (~2% of CI time) · across 27 patterns
Estimated from wall-clock time · public repos pay $0, $ = private-repo equivalent
See all 27 fixes, each with the exact YAML ↓
3.6%
failure rate, 30d
0m
avg time to recover from a failure
45 workflows · 500 runs (16.7/day) · 2,563 CI-min (wall-clock) · ≈$15 at private-repo rates (30d)

Where the minutes go (30d)

Integration Test [Kubernetes]~449 min · 33 runs
Test And Build~345 min · 36 runs
Create and publish Docker images~289 min · 17 runs
or track on every push →

Waste detected

Biggest wins first, each with the exact config fix.

No concurrency control · Content Moderation

~21 min/mo

Add a `concurrency:` block keyed on branch to cancel superseded runs when devs push twice quickly.

concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

Full guide: how to fix this →

No concurrency control · Issue Manager

~17 min/mo

Add a `concurrency:` block keyed on branch to cancel superseded runs when devs push twice quickly.

concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

Full guide: how to fix this →

No concurrency control · Owner Notification

~2 min/mo

Add a `concurrency:` block keyed on branch to cancel superseded runs when devs push twice quickly.

concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

Full guide: how to fix this →

Also found: 24 more config fixes with negligible recoverable minutes — timeouts, retention, path filters

Large build matrix · Create and publish Docker images

~0 min/mo

The matrix expands to many parallel jobs per run, multiplying billable minutes, often more combos than you need. Trim the axes, or use `include:` to list only the combinations that matter.

strategy:
  fail-fast: true
  matrix:
    include:
      - { os: ubuntu-latest, node: 20 }
      - { os: ubuntu-latest, node: 22 }

Full guide: how to fix this →

cancel-in-progress on a deploy workflow · Create and publish Docker images

~0 min/mo

This looks like a deploy/release workflow with `cancel-in-progress: true`, which cancels queued runs too — a queued deploy is silently dropped when a newer run arrives. Use a group-only `concurrency:` block on deploys.

concurrency:
  group: ${{ github.workflow }}
  # queued deploys run in order; none are dropped

Full guide: how to fix this →

ARM builds emulated under QEMU · Create and publish Docker images

~0 min/mo

ARM targets build under QEMU emulation on an x86 runner — often 10-100× slower than native. GitHub's `ubuntu-24.04-arm` runners are free for public repos; build the ARM leg natively and merge with a manifest step.

jobs:
  build-arm:
    runs-on: ubuntu-24.04-arm  # native, no QEMU — free for public repos

Full guide: how to fix this →

No job timeout · Owner Notification

~0 min/mo

No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.

jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

Full guide: how to fix this →

No job timeout · Pre-commit

~0 min/mo

No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.

jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

Full guide: how to fix this →

No job timeout · Test And Build

~0 min/mo

No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.

jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

Full guide: how to fix this →

No path filters on triggers · Test And Build

~0 min/mo

Runs on every push/PR with no `paths:` filter, so docs-only changes still trigger full CI. Add a `paths:` filter if that's common.

on:
  pull_request:
    paths:
      - 'src/**'
      - 'package.json'

Full guide: how to fix this →

Scheduled at the top of the hour · Test And Build

~0 min/mo

The schedule fires at minute :00 — GitHub's peak window, where scheduled runs get delayed or skipped. Shift to any other minute for the same cadence with less contention.

on:
  schedule:
    - cron: '33 2 * * *'  # was '0 2 * * *' — any non-:00 minute avoids the herd

Full guide: how to fix this →

No job timeout · Issue Manager

~0 min/mo

No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.

jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

Full guide: how to fix this →

System packages reinstalled every run · Integration Test [Kubernetes]

~0 min/mo

System packages install from the network on every run with no cache. Cache them (cache-apt-pkgs-action) or check whether the runner image already has the tool.

- uses: awalsh128/cache-apt-pkgs-action@latest
  with:
    packages: <your packages>
    version: 1.0

Full guide: how to fix this →

No job timeout · Content Moderation

~0 min/mo

No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.

jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

Full guide: how to fix this →

No path filters on triggers · Content Moderation

~0 min/mo

Runs on every push/PR with no `paths:` filter, so docs-only changes still trigger full CI. Add a `paths:` filter if that's common.

on:
  pull_request:
    paths:
      - 'src/**'
      - 'package.json'

Full guide: how to fix this →

Full-history clone (fetch-depth: 0) · Performance Tests

~0 min/mo

checkout fetches the entire git history every run (`fetch-depth: 0`) and nothing in the workflow appears to read it. Remove the line — the default shallow clone is much faster on big repos. Keep it if a step genuinely needs history.

- uses: actions/checkout@v4
  # fetch-depth: 0 removed — default shallow clone is enough here

Full guide: how to fix this →

No job timeout · Integration Test [Dashboard]

~0 min/mo

No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.

jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

Full guide: how to fix this →

No path filters on triggers · Integration Test [Dashboard]

~0 min/mo

Runs on every push/PR with no `paths:` filter, so docs-only changes still trigger full CI. Add a `paths:` filter if that's common.

on:
  pull_request:
    paths:
      - 'src/**'
      - 'package.json'

Full guide: how to fix this →

No job timeout · Publish Python Package to PyPI

~0 min/mo

No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.

jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

Full guide: how to fix this →

cancel-in-progress on a deploy workflow · Publish Python Package to PyPI

~0 min/mo

This looks like a deploy/release workflow with `cancel-in-progress: true`, which cancels queued runs too — a queued deploy is silently dropped when a newer run arrives. Use a group-only `concurrency:` block on deploys.

concurrency:
  group: ${{ github.workflow }}
  # queued deploys run in order; none are dropped

Full guide: how to fix this →

ARM builds emulated under QEMU · Operator CI

~0 min/mo

ARM targets build under QEMU emulation on an x86 runner — often 10-100× slower than native. GitHub's `ubuntu-24.04-arm` runners are free for public repos; build the ARM leg natively and merge with a manifest step.

jobs:
  build-arm:
    runs-on: ubuntu-24.04-arm  # native, no QEMU — free for public repos

Full guide: how to fix this →

No job timeout · Supply Chain Security Scan

~0 min/mo

No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.

jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

Full guide: how to fix this →

Full-history clone (fetch-depth: 0) · Supply Chain Security Scan

~0 min/mo

checkout fetches the entire git history every run (`fetch-depth: 0`) and nothing in the workflow appears to read it. Remove the line — the default shallow clone is much faster on big repos. Keep it if a step genuinely needs history.

- uses: actions/checkout@v4
  # fetch-depth: 0 removed — default shallow clone is enough here

Full guide: how to fix this →

No job timeout · Check Linked Issue

~0 min/mo

No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.

jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

Full guide: how to fix this →

No path filters on triggers · Check Linked Issue

~0 min/mo

Runs on every push/PR with no `paths:` filter, so docs-only changes still trigger full CI. Add a `paths:` filter if that's common.

on:
  pull_request:
    paths:
      - 'src/**'
      - 'package.json'

Full guide: how to fix this →

No job timeout · Router Learning Eval

~0 min/mo

No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.

jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

Full guide: how to fix this →

Artifacts at default retention · Router Learning Eval

~0 min/mo

`upload-artifact` has no `retention-days`, so artifacts keep up to 90 days (storage cost). Set e.g. `retention-days: 7`.

- uses: actions/upload-artifact@v4
  with:
    name: build
    path: dist/
    retention-days: 7

Full guide: how to fix this →

Copy every fix as one block (10 snippets, annotated per workflow)
# No concurrency control — applies to: Content Moderation, Issue Manager, Owner Notification
concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

# Large build matrix — applies to: Create and publish Docker images
strategy:
  fail-fast: true
  matrix:
    include:
      - { os: ubuntu-latest, node: 20 }
      - { os: ubuntu-latest, node: 22 }

# cancel-in-progress on a deploy workflow — applies to: Create and publish Docker images, Publish Python Package to PyPI
concurrency:
  group: ${{ github.workflow }}
  # queued deploys run in order; none are dropped

# ARM builds emulated under QEMU — applies to: Create and publish Docker images, Operator CI
jobs:
  build-arm:
    runs-on: ubuntu-24.04-arm  # native, no QEMU — free for public repos

# No job timeout — applies to: Owner Notification, Pre-commit, Test And Build, Issue Manager, Content Moderation, Integration Test [Dashboard], Publish Python Package to PyPI, Supply Chain Security Scan, Check Linked Issue, Router Learning Eval
jobs:
  build:
    runs-on: ubuntu-latest
    timeout-minutes: 15

# No path filters on triggers — applies to: Test And Build, Content Moderation, Integration Test [Dashboard], Check Linked Issue
on:
  pull_request:
    paths:
      - 'src/**'
      - 'package.json'

# Scheduled at the top of the hour — applies to: Test And Build
on:
  schedule:
    - cron: '33 2 * * *'  # was '0 2 * * *' — any non-:00 minute avoids the herd

# System packages reinstalled every run — applies to: Integration Test [Kubernetes]
- uses: awalsh128/cache-apt-pkgs-action@latest
  with:
    packages: <your packages>
    version: 1.0

# Full-history clone (fetch-depth: 0) — applies to: Performance Tests, Supply Chain Security Scan
- uses: actions/checkout@v4
  # fetch-depth: 0 removed — default shallow clone is enough here

# Artifacts at default retention — applies to: Router Learning Eval
- uses: actions/upload-artifact@v4
  with:
    name: build
    path: dist/
    retention-days: 7

Each snippet is representative — merge into the named workflow files rather than pasting wholesale.

Want this on every push?

This scorecard is a one-time snapshot. Install the free GitHub App to track this repo continuously: new regressions caught as they land, trends over time, on your public and private repos. Team adds the offending commit on the PR + Slack alerts.

Install & monitor this repo →

Not ready to install? Get this report by email. No spam, unsubscribe anytime.

Share this scorecard: https://gitspider.com/scan/vllm-project/semantic-router
Add the badge to your README

Live CI-health badge → GitSpider badge

[![GitSpider](https://gitspider.com/badge/vllm-project/semantic-router.svg)](https://gitspider.com/scan/vllm-project/semantic-router)