Files
hyperframes/.github/workflows/ci.yml
T
Miguel Ángel f24ff83834 feat(catalog): add the seven 3D-motion pieces as installable registry blocks (#4056)
* feat(catalog): add the seven 3D-motion pieces as installable registry blocks

Each piece ships as a registry block with a CDN poster and preview, a live
catalog detail page and its source. Scripts travel inside the catalog payload
(shared vendor libraries fetched once, item scripts inline, the one HDR as a
data URI) because the docs host only serves fonts, images and JSON.

NO_ESSAY_COMMENTS_OK: the long comment blocks are in vendored libraries and
the ported frost source under registry/blocks, not hand-written here.
HF_MAX_NONLFS_KB=6000: four hosted texture PNGs under docs/public/catalog/items
are runtime assets the docs host must serve as files.
fallow skipped locally: it audits against origin/main and this branch is
stacked on an older main.

* chore(catalog): regenerate gallery data and nav after rebasing onto main

* fix(catalog): read the install manifest through the bounded project reader

* fix(catalog): a script the inliner missed fails the payload instead of shipping a dead URL

* refactor(catalog): the leftover-script check reuses the reference finder

* fix(catalog): frost and glass textures load from the CDN and the iteration presets are gone

* chore(catalog): regenerate the frost and glass payloads and nav

* fix(catalog): frost and glass textures are hosted through the registry url field, not hardcoded

* chore(catalog): regenerate the frost and glass payloads against the hosted texture URLs

* fix(catalog): the hosted-reference rewrite refuses a bare name it cannot safely replace

* ci(catalog): verify changed catalog payloads in headless Chrome and check the docs catalog

* feat(catalog): every item page follows the shared layout with a Tune panel and tabs

* fix(catalog): let the texture-mask item host its masks so its tile plays live

* chore(catalog): regenerate item pages, gallery data and the texture-mask payload

* chore(catalog): regenerate item pages after merging main

* fix(catalog): declared-only variables get the Tune panel, Code and Docs tabs capped

* feat(catalog): give the blend-difference caption a demo scene so its tile plays live

* chore(catalog): live payload for the blend-difference caption

* perf(catalog): heavy 3d pieces compile and draw before ready, not on the first visible frame

* feat(catalog): the avatar promo card hosts its videos, audio and fonts on the CDN

* feat(catalog): the avatar promo card tile plays live from its CDN-hosted media

* chore(catalog): the avatar promo card plays live, no allowlist left

* chore(catalog): shorten two comment blocks

* chore(catalog): trim first-party comment blocks to the comment-check limits

* chore(catalog): shorten two frost source comments

* fix(catalog): the Docs tab renders as prose again, and the detail page honours reduced motion

Frame-mode pages carry no prose wrapper, and the detail component sat inside
not-prose, so the Docs slot's headings, paragraphs, links and lists had no
styling at all. The slots now carry their own prose classes and the chrome
around them keeps not-prose.

The recorded clip also gets the reduced-motion guard the docs-snippet check
requires, and the generator reads stability from the source metadata type so
the scripts typecheck passes.

* fix(catalog): docs-tab paragraphs render as blocks

Mintlify emits paragraphs as inline spans, so two paragraphs in the Docs slot ran together.

* fix(ci): drop the duplicate docs_catalog job output

git's line-based auto-merge silently duplicated the outputs key when
main and this branch each added it on non-adjacent lines during the
rebase; oxfmt's yaml parser rejects the repeated map key.

* fix(catalog): close four CodeQL alerts in the frost source

* chore(catalog): allowlist the avatar-promo-card CDN asset deletions, format sync-docs-catalog

check-no-main-deletions.mjs was rejecting 11 local mp4/font files this branch
already deleted in favor of CDN hosting; scripts/sync-docs-catalog.mjs had one
line over the 100-col print width from an earlier edit in this branch.

* fix(catalog): detail tabs use docs prose spacing, Install first; previews skip no-WebGPU runners

* fix(catalog): the docs tab scrolls with the page instead of a capped inner box

* refactor(catalog): split the payload asset and verify scripts into single-purpose helpers

* refactor(catalog): split the preview generator into single-purpose helpers

* ci(catalog): fail when the committed catalog differs from generator output

The docs host serves the committed catalog and nothing regenerates it, so a stale
copy shipped silently. The catalog payloads job now generates into a temp dir and
compares it with the committed tree.

* chore(catalog): regenerate docs/public/catalog with the refactored scripts

One full regeneration from an empty output tree with the current scripts. The tree grows
from 58.4 MB to 65.1 MB. Two hosted assets are about 2 MB each (an avatar png and a wav);
the same bytes are already committed under assets/, so HF_MAX_NONLFS_KB is raised for
this commit only.

* fix(catalog): run the catalog script tests under the runner they are written for

The rebase put three node:test files in the vitest half of test:scripts, where they report
"No test suite found". They now sit in the node --test list.

* fix(catalog): a WebGPU item without an adapter shows its recorded clip and says why

* ci(catalog): the drift check runs from a committed fetch mirror with no network

Generation now serves every font and CDN script fetch from scripts/catalog-fetch-mirror and fails
when a fetch is not in it, even if the caller swallowed the error. CATALOG_FETCH_MIRROR=record
refreshes the mirror from the live network. The docs_catalog filter also watches packages/core,
packages/producer and the mirror, and the 13 re-hashed font files are listed in ALLOWED_DELETIONS.

* chore(catalog): clean up the temp font cache and sort the mirror index without a locale

* fix(catalog): hide Tune when the clip stands in, time out the adapter request, share one check

* fix(catalog): the adapter probe survives a sync throw and a hang; test the clip guard

* fix(catalog): keep the adapter probe inside the component, Mintlify evaluates only exports

* fix(catalog): run the detail and page generator tests in CI; no Tune flash while probing WebGPU

* docs(catalog): credit vendored libraries and make each SKILL.md list what the block installs

CREDITS.md names every vendored dependency with its upstream and licence, and flags the two leaf
textures whose origin is still to be confirmed. Each SKILL.md Files table now comes from the block
manifest and its install and render paths point at compositions/<name>/. A test checks that every
path a SKILL.md names is a manifest file that exists.

* fix(catalog): a declared WebGPU item passes verify by its manifest tag, not by console wording

verify-catalog-payloads forgave a missing adapter only by matching the markup and the message, which
library-reported errors and widgets never satisfy. The webgpu tag in the item manifest is now the
owner: a tagged item ignores its page and console errors, an untagged one gets no forgiveness, and a
marker payload with no html is listed instead of crashing the run.

* docs(catalog): correct the credits per block and guard the mount and render paths

CREDITS.md states the three.js version and vendored-or-CDN status per block, adds d3-delaunay,
Clipper on the CDN and the Google Fonts loads, and lists every texture with no known origin.
The SKILL.md test now also checks the mount and render paths against the manifest composition.

* test(catalog): read each block manifest once in the SKILL.md file test

The two checks shared a copied manifest read, which fallow flagged as a clone with a high score.

* fix(catalog): commit the hosted item files a gitignore rule kept out, and check git not the disk

The compositions/ and *-ad/ ignore rules matched folder names under docs/public/catalog/items,
so 33 hosted fonts, textures and marks were generated but never committed and would 404 on
the docs host. The rules now have a negation for that tree and the files are added.

The drift check read the checkout, where git-ignored files still exist, so it passed locally
and failed only in CI. It now unpacks HEAD with git archive and compares that.

* fix(catalog): a tile keeps its poster until the composition has painted

* fix(catalog): a tagged WebGPU item forgives only the adapter-absent error, and test the wiring

The manifest webgpu tag now decides whether an item may lack an adapter, and the error wording
decides which lines are that absence. Any other page or console error on the item, and every
failed request, still fails. The per-item decision is one exported function the tests drive,
marker payloads are parsed by a tested helper, and the SKILL.md variable counts are checked
against their tables (cuboid-carousel said forty-nine for forty).

* fix(catalog): the pinned group's posters load eagerly at high priority

* fix(catalog): type the tile reveal test so the scripts typecheck passes

* fix(catalog): stop the adapter signature matching an unrelated failed adapter request

failed to request .*adapter also matched errors about other adapters. It now matches only a
request for the WebGPU adapter, so a tagged item cannot hide an unrelated failure behind it.

* test(catalog): a tagged item still fails on an error about some other adapter

* feat(catalog): the install command sits above the preview, as on the live catalog

* fix(catalog): find the install slot through Mintlify's child wrapper

* fix(docs): catalog copy buttons keep their size when the label swaps to Copied

* refactor(docs): name the shown copy label once in the catalog copy button

* fix(docs): primary catalog button keeps a readable label on hover and after copy

* fix(docs): inset the catalog tab strip and tab content to match the boxed panels
2026-09-19 16:53:29 -04:00

963 lines
38 KiB
YAML

name: CI
permissions:
contents: read
pull-requests: read
# Suppress hyperframes CLI telemetry from HeyGen's own CI runs.
# External users' CI continues to emit telemetry unless they set this themselves.
env:
HYPERFRAMES_NO_TELEMETRY: "1"
on:
pull_request:
# `edited` is required so the workflow re-fires when a PR's base ref is
# set back to `main` after a Graphite stack restack momentarily flips
# the base off of `main`. Without it, `pull_request` triggers are not
# re-evaluated on `base_ref_changed`, leaving required checks skipped
# for that head SHA forever.
types: [opened, synchronize, reopened, edited]
branches: [main]
push:
branches: [main]
concurrency:
group: ci-${{ github.ref }}
cancel-in-progress: true
jobs:
changes:
name: Detect changes
runs-on: ubuntu-latest
timeout-minutes: 2
outputs:
code: ${{ steps.filter.outputs.code }}
docs_catalog: ${{ steps.filter.outputs.docs_catalog }}
cli: ${{ steps.filter.outputs.cli }}
skills: ${{ steps.filter.outputs.skills }}
catalog_index: ${{ steps.filter.outputs.catalog_index }}
codex_plugin: ${{ steps.filter.outputs.codex_plugin }}
gcp_beginframe: ${{ steps.filter.outputs.gcp_beginframe }}
studio: ${{ steps.filter.outputs.studio }}
steps:
# Force git-based change detection instead of the pull_request REST API.
# The API path can fail the whole workflow on transient listFiles
# timeouts before any real CI work starts.
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
fetch-depth: 0
- name: Reject accidental file deletions
if: github.event_name == 'pull_request'
run: node scripts/check-no-main-deletions.mjs --base origin/main
- uses: dorny/paths-filter@fbd0ab8f3e69293af611ebaee6363fc25e6d187d # v4
id: filter
with:
token: ""
filters: |
catalog_index:
- "registry/registry.json"
- "registry/catalog-artifact/**"
- "scripts/catalog/check-artifact-coverage.ts"
docs_catalog:
- "docs/**"
- "registry/**"
- "scripts/*catalog*"
- "scripts/catalog-fetch-mirror/**"
- "scripts/docs-catalog-shared.mjs"
- "packages/core/**"
- "packages/producer/**"
- "package.json"
- ".github/workflows/ci.yml"
code:
- "packages/**"
- "scripts/**"
- "package.json"
- "bun.lock"
- "tsconfig*.json"
- "Dockerfile*"
- ".github/workflows/**"
cli:
- "packages/cli/**"
- "package.json"
- "bun.lock"
- ".github/workflows/ci.yml"
skills:
- "skills/**"
- ".claude/skills/**"
- ".agents/skills/**"
- "skills-manifest.json"
- "scripts/lint-skills.ts"
- "scripts/check-skill-mirror.mjs"
- "package.json"
- ".github/workflows/ci.yml"
codex_plugin:
- ".codex-plugin/**"
- "assets/**"
- "skills/**"
- "scripts/package-codex-plugin.mjs"
- "package.json"
- ".github/workflows/ci.yml"
studio:
- "packages/studio/**"
- "packages/core/**"
- "packages/studio-server/**"
- "bun.lock"
- ".github/workflows/ci.yml"
gcp_beginframe:
- "packages/gcp-cloud-run/Dockerfile"
- "packages/aws-lambda/scripts/probe-beginframe.ts"
- "packages/engine/src/services/browserManager.ts"
- "package.json"
- "bun.lock"
- ".github/workflows/ci.yml"
gcp-beginframe-contract:
name: GCP BeginFrame image contract
needs: changes
if: needs.changes.outputs.gcp_beginframe == 'true'
runs-on: ubuntu-latest
timeout-minutes: 15
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
- uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # v3
- uses: docker/build-push-action@263435318d21b8e681c14492fe198d362a7d2c83 # v6
with:
context: .
file: packages/gcp-cloud-run/Dockerfile
target: beginframe-contract
push: false
cache-from: type=gha,scope=gcp-beginframe-contract
cache-to: type=gha,mode=max,scope=gcp-beginframe-contract
build:
name: Build
needs: changes
if: needs.changes.outputs.code == 'true'
runs-on: ubuntu-latest
timeout-minutes: 10
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- run: corepack enable
- run: corepack prepare pnpm@10.17.1 --activate
- uses: ./.github/actions/prepare-ffmpeg-bin
- run: bash scripts/ci/install-workspace-dependencies.sh
- run: bun run build
- run: bun run verify:packed-manifests
lint:
name: Lint
needs: changes
if: needs.changes.outputs.code == 'true' || needs.changes.outputs.docs_catalog == 'true'
runs-on: ubuntu-latest
timeout-minutes: 5
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
fetch-depth: 0
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- uses: ./.github/actions/prepare-ffmpeg-bin
- run: bash scripts/ci/install-workspace-dependencies.sh
- run: bun run lint
- run: bun run check:docs-catalog
# Loads every catalog payload a PR changes in headless Chrome. A static scan of the
# markup cannot see an asset a script fetches, and the docs host serves only fonts,
# images and JSON, so a payload that runs locally can 404 there.
catalog-payloads:
name: Catalog payloads
needs: changes
if: needs.changes.outputs.docs_catalog == 'true'
runs-on: ubuntu-latest
timeout-minutes: 15
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
fetch-depth: 0
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- run: bash scripts/ci/install-workspace-dependencies.sh
- run: bun run --filter '@hyperframes/{parsers,lint,studio-server}' build
- run: bun run --cwd packages/core build
- run: bun run --cwd packages/core build:hyperframes-runtime
- run: bun run verify:catalog-payloads --changed origin/main
- run: bun run check:catalog-drift
# `fallow audit` runs dead-code + complexity + duplication analysis scoped to
# the changed files. The default `--gate new-only` means existing legacy
# findings don't fail the build — only NEW issues introduced by the PR do.
# This stops bleeding while letting incremental cleanup land separately.
#
# On findings, the job posts (or updates) a sticky comment on the PR so
# reviewers see the full list inline instead of digging through CI logs.
fallow:
name: Fallow audit
needs: changes
if: needs.changes.outputs.code == 'true' && github.event_name == 'pull_request'
runs-on: ubuntu-latest
timeout-minutes: 5
# Scope write access to this single job — the rest of `ci.yml` keeps the
# workflow-level `pull-requests: read` default so build / lint / test
# tokens can't post or modify PR comments. Job-level permissions override
# the workflow block.
permissions:
contents: read
pull-requests: write
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
# Full history so `--base origin/main` can diff against the merge
# base on stacked PRs, not just the shallow tip.
fetch-depth: 0
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- uses: ./.github/actions/prepare-ffmpeg-bin
- run: bash scripts/ci/install-workspace-dependencies.sh
- name: Run fallow audit
id: audit
# `bun install` above made `bunx fallow` resolve from node_modules, so
# we don't re-download fallow each run. The script disables `errexit`
# so the audit's non-zero exit (on findings) doesn't abort before we
# write the exit code to the step output. The size check guards
# against fallow crashing before producing markdown (e.g. transient
# parse failure) — without it we'd post a blank sticky comment.
run: |
set +e
bunx fallow audit --base origin/main --fail-on-issues \
--format pr-comment-github \
> /tmp/fallow-comment.md
echo "exit_code=$?" >> "$GITHUB_OUTPUT"
if [ ! -s /tmp/fallow-comment.md ]; then
echo "fallow produced no output — see the job logs above." > /tmp/fallow-comment.md
fi
- name: Post sticky comment (findings)
if: steps.audit.outputs.exit_code != '0'
# Fork PRs run with a read-only GITHUB_TOKEN regardless of the
# workflow's `permissions:` block, so the comment post will fail on
# forks. Don't fail the whole job — the audit gate below still fires.
continue-on-error: true
uses: marocchino/sticky-pull-request-comment@773744901bac0e8cbb5a0dc842800d45e9b2b405 # v2.9.4
with:
# `header` matches fallow's built-in `<!-- fallow-id: fallow-results -->`
# sentinel so subsequent runs update the same comment.
header: fallow-results
path: /tmp/fallow-comment.md
- name: Remove stale sticky comment (clean run)
if: steps.audit.outputs.exit_code == '0'
continue-on-error: true
uses: marocchino/sticky-pull-request-comment@773744901bac0e8cbb5a0dc842800d45e9b2b405 # v2.9.4
with:
header: fallow-results
delete: true
- name: Fail if audit found issues
if: steps.audit.outputs.exit_code != '0'
run: |
echo "::error::Fallow audit found new issues — see the PR comment above for details."
exit 1
format:
name: Format
runs-on: ubuntu-latest
timeout-minutes: 5
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- uses: ./.github/actions/prepare-ffmpeg-bin
- run: bash scripts/ci/install-workspace-dependencies.sh
- run: bun run format:check
typecheck:
name: Typecheck
needs: changes
if: needs.changes.outputs.code == 'true'
runs-on: ubuntu-latest
timeout-minutes: 10
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- uses: ./.github/actions/prepare-ffmpeg-bin
- run: bash scripts/ci/install-workspace-dependencies.sh
- run: bun run build
- run: bun run --filter '*' typecheck
- run: bun run typecheck:scripts
test:
name: Test
needs: [changes, producer-source-tests]
# Keep the existing required `Test` context authoritative for producer
# failures too. The dedicated producer matrix remains parallel and legible,
# while this job fails closed if either lane fails or is cancelled.
if: always() && needs.changes.outputs.code == 'true'
runs-on: ubuntu-latest
timeout-minutes: 10
steps:
- name: Require producer source tests
if: needs.producer-source-tests.result != 'success'
run: |
echo "::error::Producer unit/integration tests did not succeed."
exit 1
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- uses: ./.github/actions/prepare-ffmpeg-bin
- run: bash scripts/ci/install-workspace-dependencies.sh
- run: bun run --filter '@hyperframes/{parsers,lint,studio-server}' build
- run: bun run --cwd packages/core build
- run: bun run --cwd packages/core build:hyperframes-runtime
# After the builds: the preview scripts drive the producer, which imports
# @hyperframes/core from dist, so running these first fails on a missing
# module rather than on anything the test is actually checking.
- run: bun run test:scripts
- run: bun run --filter '!@hyperframes/producer' test
producer-source-tests:
name: "Producer: ${{ matrix.lane }} tests"
needs: changes
if: needs.changes.outputs.code == 'true'
runs-on: ubuntu-latest
timeout-minutes: 20
strategy:
fail-fast: false
matrix:
lane: [unit, integration]
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- name: Install FFmpeg for integration tests
if: matrix.lane == 'integration'
uses: ./.github/actions/install-ffmpeg-linux
- uses: ./.github/actions/prepare-ffmpeg-bin
- run: bash scripts/ci/install-workspace-dependencies.sh
- run: bun run --filter '@hyperframes/{parsers,lint,studio-server}' build
- run: bun run --cwd packages/core build
- run: bun run --filter @hyperframes/engine build
- run: bun run producer:test:${{ matrix.lane }}
# Tests under skills/**/*.test.mjs are bare `node --test` files with only
# `node:` built-in imports. They aren't part of any workspace package, and
# the main `Test` job's `code` path filter excludes `skills/**`, so without
# this dedicated job they'd never run in CI. Examples:
# * skills/media-use/scripts/resolve.test.mjs
# * skills/media-use/scripts/lib/manifest.test.mjs
# Several of these are regression guards (e.g. shell-injection cases), so
# the whole point is that they fire on PRs that touch skills/.
test-skills:
name: "Test: skills"
needs: changes
if: needs.changes.outputs.skills == 'true'
runs-on: ubuntu-latest
timeout-minutes: 5
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- name: Discover and run skills tests
# We expand the test list via bash so the job fails loudly when the
# matcher comes back empty, rather than silently no-op'ing (which
# would defeat the whole point of this job).
run: |
set -euo pipefail
mapfile -t SKILLS_TESTS < <(find skills -type f -name "*.test.mjs" | sort)
if [ "${#SKILLS_TESTS[@]}" -eq 0 ]; then
echo "::error::No skills/**/*.test.mjs files found. Did the layout change?"
exit 1
fi
printf 'Running %d skills test file(s):\n' "${#SKILLS_TESTS[@]}"
printf ' * %s\n' "${SKILLS_TESTS[@]}"
node --test "${SKILLS_TESTS[@]}"
test-skill-bootstrap-windows:
name: "Test: skill bootstrap (Windows, Node 24)"
needs: changes
if: needs.changes.outputs.skills == 'true'
runs-on: windows-latest
timeout-minutes: 5
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 24
- name: Exercise npm-cli bootstrap command
run: >-
node --test
skills/hyperframes-animation/scripts/package-loader.test.mjs
skills/hyperframes-creative/scripts/package-loader.test.mjs
# Guards that skills-manifest.json (the published freshness fingerprint read
# by `hyperframes skills check`) was regenerated when a skill changed. Runs
# `gen:skills-manifest --check`, which compares per-skill content hashes; the
# manifest carries no version/timestamp, so it only fails on real content
# drift. bun runs the TS script directly, no install needed.
catalog-index-coverage:
name: "Catalog: search index covers the registry"
needs: changes
if: needs.changes.outputs.catalog_index == 'true'
runs-on: ubuntu-latest
timeout-minutes: 3
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
# Comparing two name lists needs neither the embedding model nor a
# network call, so this stays a seconds-long gate. Regenerating the
# vectors does need the model, which is why CI checks rather than fixes.
- name: Verify every searchable registry item has a vector
run: bun scripts/catalog/check-artifact-coverage.ts
skills-manifest:
name: "Skills: manifest in sync"
needs: changes
if: needs.changes.outputs.skills == 'true'
runs-on: ubuntu-latest
timeout-minutes: 3
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- name: Verify skills-manifest.json matches skills/
run: bun packages/cli/scripts/gen-skills-manifest.ts --check
# Repo-native project skills live under .claude/skills/ (Claude Code) and
# .agents/skills/ (Codex CLI). Runs the skill validator against all three
# locations (skills/, .claude/skills/, .agents/skills/) and verifies the two
# project-skill trees stay byte-identical mirrors of each other. Fires on
# any change to the skill dirs or the two validating scripts.
project-skills-check:
name: "Skills: project-native lint + mirror"
needs: changes
if: needs.changes.outputs.skills == 'true'
runs-on: ubuntu-latest
timeout-minutes: 3
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- name: Lint SKILL.md files across skills/, .claude/skills/, .agents/skills/
run: bun scripts/lint-skills.ts
- name: Verify .claude/skills/ and .agents/skills/ are byte-identical
run: node scripts/check-skill-mirror.mjs
codex-plugin-package:
name: "Codex plugin package"
needs: changes
if: needs.changes.outputs.codex_plugin == 'true'
runs-on: ubuntu-latest
timeout-minutes: 3
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- name: Build upload-ready Codex plugin
run: node scripts/package-codex-plugin.mjs
cli-npx-shim:
name: "CLI: npx shim (${{ matrix.os }})"
needs: changes
if: needs.changes.outputs.cli == 'true'
runs-on: ${{ matrix.os }}
timeout-minutes: 10
strategy:
fail-fast: false
matrix:
os: [ubuntu-latest, macos-latest, windows-latest]
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- name: Install dependencies
if: runner.os != 'Windows'
run: bash scripts/ci/install-workspace-dependencies.sh --ignore-scripts
- name: Install dependencies
if: runner.os == 'Windows'
run: bun install --frozen-lockfile --ignore-scripts --linker=hoisted
- run: bun run --cwd packages/cli test src/utils/npxCommand.test.ts src/commands/skills.test.ts
sdk-tests:
name: "SDK: unit + contract + smoke"
needs: changes
if: needs.changes.outputs.code == 'true'
runs-on: ubuntu-latest
timeout-minutes: 5
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- run: bash scripts/ci/install-workspace-dependencies.sh
# Build workspace deps so the sdk's @hyperframes/parsers + core subpath
# imports resolve via the "node" export condition (dist) under vitest.
- run: bun run --filter '@hyperframes/parsers' build
- run: bun run --cwd packages/core build
- run: bun run --filter @hyperframes/sdk test
test-runtime-contract:
name: "Test: runtime contract"
needs: changes
if: needs.changes.outputs.code == 'true'
runs-on: ubuntu-latest
timeout-minutes: 10
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- uses: ./.github/actions/prepare-ffmpeg-bin
- run: bash scripts/ci/install-workspace-dependencies.sh
# Runtime coverage now imports core modules that consume workspace
# subpaths. Build their dist exports before Vitest resolves them.
- run: bun run --filter '@hyperframes/{parsers,lint,studio-server}' build
- run: bun run --filter @hyperframes/core test:hyperframe-runtime-ci
studio-load-smoke:
name: "Studio: load smoke"
needs: [changes]
if: needs.changes.outputs.code == 'true'
runs-on: ubuntu-latest
timeout-minutes: 5
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- uses: ./.github/actions/prepare-ffmpeg-bin
- run: bash scripts/ci/install-workspace-dependencies.sh
# Build workspace deps so the studio vite.config.ts (loaded by Node) can
# resolve @hyperframes/core and @hyperframes/studio-server via the "node"
# export condition (dist).
- run: bun run --filter '@hyperframes/{parsers,lint,studio-server}' build
- run: bun run --cwd packages/core build
- run: bun run --cwd packages/core build:hyperframes-runtime
- name: Start studio and check for runtime errors
run: |
# Start the studio Vite dev server (fast — no bundle step)
bun run --filter '@hyperframes/studio' dev -- --port 5199 &
SERVER_PID=$!
# Wait for the server to be ready (up to 20s)
for i in $(seq 1 40); do
if curl -sf http://localhost:5199/ >/dev/null 2>&1; then break; fi
sleep 0.5
done
if ! curl -sf http://localhost:5199/ >/dev/null 2>&1; then
echo "FAIL: studio dev server did not start"
kill $SERVER_PID 2>/dev/null || true
exit 1
fi
# Load the studio in headless Chrome with API mocking to trigger
# the full splash→main transition (catches hooks-after-early-return bugs)
node scripts/studio-runtime-smoke.mjs http://localhost:5199/#project=smoke-test
kill $SERVER_PID 2>/dev/null || true
studio-timeline-viewport:
name: "Studio: timeline viewport gate"
needs: [changes]
if: needs.changes.outputs.studio == 'true'
runs-on: ubuntu-latest
timeout-minutes: 12
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- run: bash scripts/ci/install-workspace-dependencies.sh
# Same reason as studio-load-smoke: vite.config.ts is loaded by Node and
# resolves the workspace packages through their "node" export condition.
- run: bun run --filter '@hyperframes/{parsers,lint,studio-server}' build
- run: bun run --cwd packages/core build
- run: bun run --cwd packages/core build:hyperframes-runtime
- name: Install the fixture as a Studio project
# Studio resolves projects from packages/studio/data/projects, which is
# gitignored. The fixture composition is tracked under tests/e2e, so
# copy it into place rather than committing a project directory.
run: |
mkdir -p packages/studio/data/projects
cp -R packages/studio/tests/e2e/fixtures/timeline-virtualization \
packages/studio/data/projects/timeline-virtualization
- name: Run both arms of the timeline viewport gate
run: |
set -euo pipefail
SERVER_PID=""
stop_server() {
if [[ -n "$SERVER_PID" ]]; then
kill "$SERVER_PID" 2>/dev/null || true
wait "$SERVER_PID" 2>/dev/null || true
SERVER_PID=""
fi
}
wait_for_server() {
local port="$1"
for i in $(seq 1 60); do
if curl -sf "http://localhost:${port}/" >/dev/null 2>&1; then return 0; fi
sleep 1
done
echo "FAIL: studio dev server did not start on port ${port}"
return 1
}
trap stop_server EXIT
# Run one server at a time so the measured browser never competes with
# a second Vite module graph on the shared runner. The development
# server supplies the fixture API; production React matches shipped
# rendering behavior, and the gate asserts that runtime before timing.
NODE_ENV=production \
bun run --cwd packages/studio dev -- --port 5313 --strictPort &
SERVER_PID=$!
DEFAULT_STATUS=0
if wait_for_server 5313; then
STUDIO_URL="http://localhost:5313/#project/timeline-virtualization" \
TIMELINE_ROW_VIRTUALIZATION=on \
TIMELINE_ELEMENT_COUNT=50000 \
TIMELINE_TIER=ci \
node packages/studio/tests/e2e/timeline-virtualization.mjs \
| tee /tmp/timeline-gate-default.json \
|| DEFAULT_STATUS=$?
else
DEFAULT_STATUS=1
fi
stop_server
NODE_ENV=production \
VITE_STUDIO_TIMELINE_ROW_VIRTUALIZATION_ENABLED=0 \
bun run --cwd packages/studio dev -- --port 5314 --strictPort &
SERVER_PID=$!
DISABLED_STATUS=0
if wait_for_server 5314; then
STUDIO_URL="http://localhost:5314/#project/timeline-virtualization" \
TIMELINE_ROW_VIRTUALIZATION=off \
TIMELINE_ELEMENT_COUNT=1000 \
TIMELINE_TIER=ci \
node packages/studio/tests/e2e/timeline-virtualization.mjs \
| tee /tmp/timeline-gate-disabled.json \
|| DISABLED_STATUS=$?
else
DISABLED_STATUS=1
fi
stop_server
{
echo "### Timeline viewport gate"
echo "- Default arm exit: ${DEFAULT_STATUS}"
echo "- Explicitly disabled arm exit: ${DISABLED_STATUS}"
} >> "$GITHUB_STEP_SUMMARY"
if (( DEFAULT_STATUS != 0 || DISABLED_STATUS != 0 )); then
echo "FAIL: default=${DEFAULT_STATUS}, disabled=${DISABLED_STATUS}"
exit 1
fi
- name: Upload gate evidence
# The gate's whole output is machine-readable evidence, and a red run is
# exactly when someone needs to read it. Keep it on failure too.
if: always()
uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4
with:
name: timeline-viewport-gate-evidence
path: /tmp/timeline-gate-*.json
if-no-files-found: error
smoke-global-install:
name: "Smoke: global install"
needs: [changes, build]
if: needs.changes.outputs.code == 'true'
runs-on: ubuntu-latest
timeout-minutes: 10
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- uses: ./.github/actions/prepare-ffmpeg-bin
- run: bash scripts/ci/install-workspace-dependencies.sh
- run: bun run build
# Pack the CLI as a tarball (simulates what `npm publish` produces)
- name: Pack CLI tarball
run: cd packages/cli && npm pack
# Install globally using --prefix to avoid sudo
- name: Install globally via npm
run: npm install -g --prefix /tmp/hf-smoke ./packages/cli/hyperframes-cli-*.tgz
# Scaffold a blank project
- name: Init blank project
run: |
export PATH="/tmp/hf-smoke/bin:$PATH"
mkdir /tmp/hf-project && cd /tmp/hf-project
hyperframes init test-project --example blank
# Start preview, probe the runtime endpoint, assert no esbuild errors
- name: Smoke-test preview server
run: |
export PATH="/tmp/hf-smoke/bin:$PATH"
cd /tmp/hf-project/test-project
# Start the preview server in the background; capture stderr
CI=true hyperframes preview --port 3099 2>/tmp/hf-stderr.log &
SERVER_PID=$!
# Wait for the server to be ready (up to 15 s)
for i in $(seq 1 30); do
if curl -sf http://localhost:3099/ >/dev/null 2>&1; then
break
fi
sleep 0.5
done
# Probe the runtime JS endpoint
BODY=$(curl -sf http://localhost:3099/api/runtime.js | head -c 200 || true)
if [ -z "$BODY" ]; then
echo "FAIL: /api/runtime.js returned empty response"
kill $SERVER_PID 2>/dev/null || true
cat /tmp/hf-stderr.log
exit 1
fi
kill $SERVER_PID 2>/dev/null || true
wait $SERVER_PID 2>/dev/null || true
# Assert stderr does not contain esbuild / runtime load errors
if grep -qE '✘ \[ERROR\]|Failed to load runtime' /tmp/hf-stderr.log; then
echo "FAIL: preview emitted runtime errors:"
cat /tmp/hf-stderr.log
exit 1
fi
echo "PASS: global install smoke test succeeded"
cli-smoke-required:
name: "CLI smoke (required)"
needs: changes
if: needs.changes.outputs.code == 'true'
runs-on: ubuntu-latest
timeout-minutes: 25
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
lfs: true
- uses: oven-sh/setup-bun@0c5077e51419868618aeaa5fe8019c62421857d6 # v2
- uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4
with:
node-version: 22
- name: Install FFmpeg
uses: ./.github/actions/install-ffmpeg-linux
- uses: ./.github/actions/prepare-ffmpeg-bin
- name: Install dependencies
run: bash scripts/ci/install-workspace-dependencies.sh
- name: Build monorepo
run: bun run build
- name: Create smoke input video
run: |
set -euo pipefail
ffmpeg -hide_banner -loglevel error \
-f lavfi -i testsrc2=size=640x360:rate=30 \
-f lavfi -i sine=frequency=880:sample_rate=48000 \
-t 3 \
-c:v libx264 \
-pix_fmt yuv420p \
-c:a aac \
-shortest \
-y /tmp/hf-cli-input.mp4
test -s /tmp/hf-cli-input.mp4
- name: Smoke-test CLI from monorepo source
# init's --skip-skills flag is neutered (see init.ts); opt out of the
# GitHub skills check via this env so the smoke test stays offline/fast.
env:
HYPERFRAMES_SKIP_SKILLS: "1"
run: |
set -euo pipefail
rm -rf /tmp/hf-cli-inside
bun run --filter @hyperframes/cli dev -- init /tmp/hf-cli-inside \
--example warm-grain \
--video /tmp/hf-cli-input.mp4 \
--skip-transcribe \
--non-interactive \
--skip-skills
bun run --filter @hyperframes/cli dev -- lint /tmp/hf-cli-inside
bun run --filter @hyperframes/cli dev -- validate /tmp/hf-cli-inside --timeout 3000
bun run --filter @hyperframes/cli dev -- render /tmp/hf-cli-inside \
--quality standard \
--workers auto \
--strict \
--output /tmp/hf-cli-inside/renders/inside.mp4 2>&1 | tee /tmp/hf-cli-render.log
test -s /tmp/hf-cli-inside/renders/inside.mp4
# Belongs in this job because it needs the BUILT CLI: `shouldTrack()`
# short-circuits whenever `isDevMode()` is true, which it is for any `.ts`
# entry, so under vitest a real event and no event look identical. Sends
# nothing to PostHog — the script wraps fetch and captures the batch.
- name: Smoke-test `add` telemetry end to end
run: node scripts/ci/cli-telemetry-e2e.mjs
- name: Assert page.goto completes under 5s budget
run: |
set -euo pipefail
# Extract the longest page.goto time from render logs.
# Format: [initSession:MODE] page.goto complete (NNNms)
MAX_MS=$(grep -oP 'page\.goto complete \(\K[0-9]+' /tmp/hf-cli-render.log | sort -n | tail -1)
if [ -z "$MAX_MS" ]; then
echo "::error::No page.goto timing found in render logs — log format may have changed. Update the grep pattern."
exit 1
fi
echo "Slowest page.goto: ${MAX_MS}ms (budget: 5000ms)"
if [ "$MAX_MS" -gt 5000 ]; then
echo "::error::page.goto took ${MAX_MS}ms — exceeds 5s budget. Possible Proxy/stub regression."
exit 1
fi
- name: Pack CLI tarball
run: |
set -euo pipefail
mkdir -p /tmp/hf-cli-pack
cd packages/cli
PACKED_TARBALL="$(npm pack --pack-destination /tmp/hf-cli-pack | tail -n 1)"
test -n "$PACKED_TARBALL"
test -f "/tmp/hf-cli-pack/$PACKED_TARBALL"
echo "HF_CLI_TARBALL=/tmp/hf-cli-pack/$PACKED_TARBALL" >> "$GITHUB_ENV"
- name: Install packed CLI outside monorepo
run: |
set -euo pipefail
npm install -g --prefix /tmp/hf-cli-global "$HF_CLI_TARBALL"
- name: Smoke-test packed CLI outside monorepo
# init's --skip-skills flag is neutered (see init.ts); opt out of the
# GitHub skills check via this env so the smoke test stays offline/fast.
env:
HYPERFRAMES_SKIP_SKILLS: "1"
run: |
set -euo pipefail
export PATH="/tmp/hf-cli-global/bin:$PATH"
rm -rf /tmp/hf-cli-outside
hyperframes init /tmp/hf-cli-outside \
--example warm-grain \
--video /tmp/hf-cli-input.mp4 \
--skip-transcribe \
--non-interactive \
--skip-skills
hyperframes lint /tmp/hf-cli-outside
hyperframes validate /tmp/hf-cli-outside --timeout 3000
hyperframes render /tmp/hf-cli-outside \
--quality standard \
--workers auto \
--strict \
--output /tmp/hf-cli-outside/renders/outside.mp4
test -s /tmp/hf-cli-outside/renders/outside.mp4
filesize:
name: File size check
runs-on: ubuntu-latest
timeout-minutes: 1
steps:
- uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5 # v4
with:
fetch-depth: 0
- name: Check file sizes (max 600 lines)
# Scoped to files THIS PR changed under packages/studio. Walking the
# whole tree blamed every unrelated PR for pre-existing offenders.
# Falls back to a full scan on push events (no base ref available)
# so the rule still guards main.
run: |
set -e
if [ -n "${{ github.base_ref }}" ]; then
mapfile -t files < <(
git diff --name-only --diff-filter=ACMR \
"origin/${{ github.base_ref }}...HEAD" -- \
'packages/studio/**/*.ts' 'packages/studio/**/*.tsx' \
| grep -vE '\.(test|spec)\.(ts|tsx)$|\.generated\.' || true
)
else
mapfile -t files < <(
find packages/studio -path '*/node_modules' -prune -o \
\( -name '*.ts' -o -name '*.tsx' \) -print \
| grep -vE '\.(test|spec)\.(ts|tsx)$|\.generated\.'
)
fi
EXIT=0
for f in "${files[@]}"; do
[ -z "$f" ] && continue
[ -f "$f" ] || continue # skip files deleted in this PR
lines=$(wc -l < "$f")
if [ "$lines" -gt 600 ]; then
echo "::error file=$f::$f has $lines lines (max 600)"
EXIT=1
fi
done
exit $EXIT
semantic-pr-title:
name: Semantic PR title
if: github.event_name == 'pull_request'
runs-on: ubuntu-latest
timeout-minutes: 5
steps:
- uses: amannn/action-semantic-pull-request@e32d7e603df1aa1ba07e981f2a23455dee596825 # v5
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
with:
types: |
feat
fix
docs
style
refactor
perf
test
build
ci
chore
revert