Files
apes-sidebar/.github/workflows/ci.yml
T

212 lines
9.0 KiB
YAML

name: CI
on:
push:
branches: [main]
pull_request:
workflow_dispatch:
# Cancel a superseded PR run (it is on a stale commit); never cancel a
# push run — it is already producing the post-merge signal. The mount job
# boots a real DSH + Chromium, so the savings matter there most.
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: ${{ github.event_name == 'pull_request' }}
jobs:
ci:
runs-on: ubuntu-latest
timeout-minutes: 20
steps:
- uses: actions/checkout@v7
with:
# Tests run git log against this repository and need the full history.
fetch-depth: 0
# pnpm version is pinned via packageManager in package.json; the action
# reads it from there (specifying both makes action-setup fail on
# purpose with ERR_PNPM_BAD_PM_VERSION).
- uses: pnpm/action-setup@v6
- uses: actions/setup-node@v7
with:
node-version: 22
cache: pnpm
# The plugin's git revert/cherry-pick spawn git without setting an
# identity (deliberate, see tests/smoke.spec.ts) and rely on ambient
# config; the runner has none by default.
- name: Configure git identity for tests
run: |
git config --global user.name "dsh-apes-sidebar-ci"
git config --global user.email "ci@dsh.invalid"
- name: Install dependencies
run: pnpm install --frozen-lockfile
- name: Typecheck
run: pnpm typecheck
- name: Lint
run: pnpm lint
- name: Test
run: pnpm test
- name: Build
run: pnpm build
# The consumer-facing declaration surface guard: a browser-only plugin
# (no @types/node, skipLibCheck: false) must type-check against the
# shipped lib/types. Runs AFTER build (it reads the built output).
- name: Check consumer type surface
run: pnpm check:consumer-types
# Windows parity lane for the pure-logic suite. The codebase carries real
# win32 branches (shell parsing, WSL path projection, the PowerShell
# installer) that mock-injection tests alone cannot validate — this lane
# runs the same typecheck / test / build / consumer-type-surface steps on
# a real Windows runtime. The plugin-mount lane (real DSH + Playwright
# Chromium) stays ubuntu-only and is deliberately NOT mirrored here.
ci-windows:
runs-on: windows-latest
timeout-minutes: 30
defaults:
run:
# The runner ships Git Bash: multi-line steps and
# scripts/check-consumer-types.sh work verbatim.
shell: bash
steps:
- uses: actions/checkout@v7
with:
# Tests run git log against this repository and need the full history.
fetch-depth: 0
# pnpm version is pinned via packageManager in package.json; the action
# reads it from there (specifying both makes action-setup fail on
# purpose with ERR_PNPM_BAD_PM_VERSION).
- uses: pnpm/action-setup@v6
- uses: actions/setup-node@v7
with:
node-version: 22
cache: pnpm
# Same rationale as the ci job: smoke tests spawn git without an
# identity and rely on ambient config, which the runner lacks.
- name: Configure git identity for tests
run: |
git config --global user.name "dsh-apes-sidebar-ci"
git config --global user.email "ci@dsh.invalid"
- name: Install dependencies
run: pnpm install --frozen-lockfile
- name: Typecheck
run: pnpm typecheck
- name: Lint
run: pnpm lint
# `test:windows` runs the whole suite in ONE fork (`--maxWorkers=1`),
# not `pnpm test`'s unbounded pool. Two distinct Windows-only failures
# motivated it, both recorded in docs/plans/2026-09-10-dsh-0.1.5-rc.2-adaptation.md:
#
# 1. Six red runs were real 5000 ms vitest timeouts in specs that
# really start a process (tests/agent-pty:71/:106,
# tests/install-powershell:105 — a cold powershell.exe start
# measured 12.1 s, tests/git.spec.ts:85 at 9.6 s). The suite now
# declares its own budgets (global testTimeout 15 s, 30 s in the
# two heaviest describes), and serial execution stops the
# process-spawning specs starving each other on a 2-core runner.
# 2. Four runs died at teardown with `[vitest-pool]: Worker forks
# emitted error / Worker exited unexpectedly` while reporting
# N-1 of N files passed and zero FAIL lines — a worker recycling
# race, not a test failure. It survived --maxWorkers=2, so the
# lane now recycles nothing at all.
#
# No retries: they would hide a real regression instead of removing the
# race. The cost is wall-clock (the Windows step runs several minutes
# against a 30-minute job budget), not signal.
- name: Test
run: pnpm test:windows
- name: Build
run: pnpm build
- name: Check consumer type surface
run: pnpm check:consumer-types
# Pack the plugin as an npm tarball, mount it into a REAL DSH instance
# through the official `dsh plugin --profile web add` channel, and render
# the DSH web UI headlessly (Playwright Chromium) to prove the plugin
# mounts without crashing the shell. Drives `pnpm test:mount`, which boots
# a keyless `dsh web` against a scratch profile and sweeps every built-in
# sidebar tab. No API key needed — the whole lane is deterministic.
plugin-mount:
runs-on: ubuntu-latest
timeout-minutes: 30
# CI runs must never report to the production telemetry endpoint baked
# into the DSH composition (same guard DSH's own CI uses).
env:
DSH_TELEMETRY_DISABLED: '1'
steps:
- uses: actions/checkout@v7
# pnpm version is pinned via packageManager in package.json; the action
# reads it from there (specifying both makes action-setup fail on
# purpose with ERR_PNPM_BAD_PM_VERSION).
- uses: pnpm/action-setup@v6
- uses: actions/setup-node@v7
with:
node-version: 22
cache: pnpm
- name: Install dependencies
run: pnpm install --frozen-lockfile
# Pin the DSH CLI the lane mounts into; the profile's bundle stack
# (dsh-base / dsh-web-app) resolves from this installation. Keep in
# step with the plugin's @deepseek-ai/* peer range when bumping.
# Write an EXACT version, never a dist-tag: when 0.1.7-rc.1 shipped, the
# npm `alpha` tag still meant 0.1.7-alpha.2, rc.1 went out on `next`, and
# `latest` was still the 0.1.5 line — `@alpha` would mount a host this
# plugin no longer supports. The 0.2.0 line repeats the pattern: its
# first candidate is `next` while `latest` is still 0.1.7-rc.2.
# The lanes speak the slash-only RPC dialect
# (tests/e2e/host-protocol.ts).
#
# The larger heap is the fix for one of the two ways this step used to
# die (runs 34490912649 and 34363627184 on the alpha line): the pinned
# version's TRANSITIVES are declared as floating `^` ranges, so while
# upstream publishes a prerelease in stages (0.1.5-rc.2 went out
# 14:43-14:57 on 2026-09-10) npm resolved a mixed peer graph —
# 3062 ERESOLVE lines, then "Ineffective mark-compacts near heap limit"
# (exit 134). The other failure mode was a hard ETARGET on a sibling
# package that appeared minutes later; pinning a version that is fully
# published (not mid-flight) is the fix for that one.
#
# Peer resolution stays ON: `--legacy-peer-deps` was tried in the first
# cut of this fix and made the install broken rather than robust — it
# skips exactly the peers a global install has to supply, and
# @deepseek-ai/dsh-app-boot requires `@deepseek-ai/cordis-plugin-group`
# (a peer, not a dependency) at boot, so the CLI died with
# ERR_MODULE_NOT_FOUND before the smoke could even start. Verified
# locally that the plain install resolves it.
- name: Install dsh CLI (pinned)
env:
NODE_OPTIONS: --max-old-space-size=4096
run: npm install -g @deepseek-ai/dsh@0.2.0-rc.1
# lib/ is gitignored; the tarball is produced from this build.
- name: Build and pack the npm tarball
run: pnpm build && pnpm pack
- name: Install Playwright Chromium
run: pnpm exec playwright install --with-deps chromium
- name: Mount + headless-render smoke
run: pnpm test:mount
# Aggregate double-mount regression: a fixture aggregate bundle mounts
# dsh-apes-sidebar under its own entry id BEFORE the plugin's bundle
# row; the plugin must back off (no duplicate /sidebar/api, boot OK).
# Uses the pinned DSH CLI installed above.
- name: Aggregate double-mount regression
env:
DSH_CMD: dsh
run: pnpm test:mount:aggregate
- name: Upload Playwright report on failure
if: failure()
uses: actions/upload-artifact@v4
with:
name: playwright-report
path: |
playwright-report/
test-results/
if-no-files-found: ignore