Files
hermes-agent/.github/workflows/install-e2e.yml
T
ethernet ea4cd375f8 ci(install-e2e): retire the bubblewrap sandbox - git redirect everywhere, macos legs live
The fake Internet (bubblewrap + slirp4netns + MITM proxy +
upload-pack shim, 883 lines across dev-sandbox.sh, stage2-run.sh,
proxy.py, ssh-shim.sh, openssl.cnf, install-update-e2e.sh) existed to
isolate install.sh's network. The GIT_CONFIG_GLOBAL insteadOf redirect
the windows driver introduced does the same job with a gitconfig file
and works on any OS, so:

* install-e2e-run.yml now runs tests/install/installer-script-e2e.sh
  directly on the bare runner - no sandbox deps, no userns sysctls -
  and takes a runner input;
* the macos matrix calls the SAME workflow on macos-latest, deleting
  install-e2e-macos-run.yml: installer-script -> installer-script /
  hermes-update flip from grey to live, app-update pairs stay TODO
  inside the shared gate;
* install.sh is no longer curl'd through a fake CA - each leg runs
  the copy from the ref a user of that version actually executed;
* scripts/dev-sandbox.sh becomes the minimal isolation sandbox from
  ab6b9492f (separate HERMES_HOME / Electron userData / app name,
  same CLI surface: --persistent, --from, --delete), keeping its
  .hermes-sandbox dir name so gitignore and docs hold;
* nix/sandbox.nix drops the bwrap/proxy closure and keeps only the
  Electron runtime LD_LIBRARY_PATH the desktop app needs.

Verified: nix build .#sandbox + smoke run (isolated HERMES_HOME
created, ephemeral cleanup), shellcheck/bash -n on both scripts,
actionlint on all three workflows, and the new driver ran the full
v0.20.2 -> HEAD hermes-update pass locally before this commit.
2026-08-11 21:20:42 -04:00

190 lines
8.0 KiB
YAML

name: Install & Update E2E
# Can a user on a released version get to this commit?
#
# The support matrix -- every {os, install-method, update-method} combination
# a user could be on -- lives in scripts/sandbox/generate-e2e-matrix.mjs.
# generate-matrix expands it against the picked release tags into one leg
# per {combination, tag}, split into one matrix job per OS:
#
# Matrix: linux the real curl|bash install one-liner, isolated by a
# git URL redirect to a local bare clone
# (install-e2e-run.yml)
# Matrix: windows the real desktop user flow: website Hermes-Setup.exe
# clicked by AutoHotkey, update via the app, Playwright
# clicking "Update now" (install-e2e-windows-run.yml)
# Matrix: macos the same OS-agnostic driver as linux, on macos-latest
# (install-e2e-run.yml; app-update pairs are TODO)
#
# Every combination is dispatched to its OS's run workflow; the run
# workflow natively skips (grey) what its driver cannot run yet -- an
# unimplemented method pair, or a starting tag that predates the surface
# under test (pick-releases annotates each tag with what its tree ships,
# e.g. whether the desktop app exists yet). Capability knowledge lives
# next to each driver, never here and never in the generator: declaring a
# method is a spec edit, implementing one is flipping the run workflow's
# gate.
#
# The starting versions are chosen at runtime from the repo's release tags
# (scripts/sandbox/pick-release-tags.sh): newest, oldest, and a spread
# between. A hardcoded list would stop covering the newest release the day
# after it ships, and would pin an "oldest" that nobody still runs.
#
# Triggers:
# * every 12 hours, so upstream drift (a new uv, a Node bump, a PyPI change)
# surfaces on a schedule rather than in someone's review cycle;
# * when a release tag is created -- the moment the set of versions users can
# update FROM changes, and the moment a broken updater would strand them;
# * manually, where you can pick the route and how many releases to sample.
#
# Deliberately NOT on pull_request: a leg takes ~11 minutes of real toolchain
# installation, and the matrix multiplies that. Updating is release-shaped work,
# so it is gated on releases and the clock instead.
on:
workflow_dispatch:
inputs:
route:
description: 'Which combinations to run. all = every OS; both/update/installer = the linux legs; windows-desktop = the windows legs.'
required: false
type: choice
default: all
options: [all, both, update, installer, windows-desktop]
tag-count:
description: 'How many release tags to sample (newest, oldest, and a spread between).'
required: false
type: string
default: '5'
schedule:
# Every 12 hours, off the hour to avoid the top-of-hour runner crunch.
- cron: '20 7,19 * * *'
push:
tags:
# Release tags only: the repo also carries backup/* and one-off tags.
- 'v[0-9]+.[0-9]+.[0-9]+'
- 'v[0-9]+.[0-9]+.[0-9]+.[0-9]+'
permissions:
contents: read
concurrency:
group: install-e2e-${{ github.ref }}
cancel-in-progress: true
jobs:
# Which released versions do we test updating FROM? Resolved once,
# annotated with what each tag's own tree supports, and shared by every
# OS's matrix so all combos cover the same set.
pick-releases:
name: Pick release tags
runs-on: ubuntu-latest
timeout-minutes: 5
outputs:
tags: ${{ steps.pick.outputs.tags }}
steps:
# This job only reads tag names and trees, so take the cheap
# checkout: no blobs (filter), no other files (sparse), but DO fetch
# tags -- they are the whole input, and the default shallow checkout
# has none.
- uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
with:
filter: blob:none
fetch-tags: true
sparse-checkout: scripts/sandbox/pick-release-tags.sh
sparse-checkout-cone-mode: false
- id: pick
run: |
set -euo pipefail
tags="$(scripts/sandbox/pick-release-tags.sh --count '${{ inputs.tag-count || 5 }}')"
echo "Testing updates from: $tags"
# Annotate each tag with what its own tree supports, so run
# workflows can natively skip surfaces the starting version does
# not have. Today: does the release ship the desktop app
# (apps/desktop, #20059)? Cheaper here -- the tags are already
# fetched -- than a probe job per leg.
enriched="$(for t in $(echo "$tags" | jq -r '.[]'); do
if git ls-tree -d "$t" apps/desktop | grep -q .; then d=true; else d=false; fi
echo "{\"ref\":\"$t\",\"desktop\":$d}"
done | jq -sc .)"
echo "Annotated: $enriched"
echo "tags=$enriched" >> "$GITHUB_OUTPUT"
# Expand the support matrix against the picked tags: one leg per
# {os, install-method, update-method, tag}, split into a matrix per OS.
generate-matrix:
name: Expand combinations
needs: pick-releases
runs-on: ubuntu-latest
timeout-minutes: 5
outputs:
linux: ${{ steps.gen.outputs.linux }}
windows: ${{ steps.gen.outputs.windows }}
macos: ${{ steps.gen.outputs.macos }}
steps:
- uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
with:
sparse-checkout: scripts/sandbox/generate-e2e-matrix.mjs
sparse-checkout-cone-mode: false
- id: gen
run: |
set -euo pipefail
matrices="$(node scripts/sandbox/generate-e2e-matrix.mjs \
--tags '${{ needs.pick-releases.outputs.tags }}')"
echo "$matrices"
for key in linux windows macos; do
echo "$key=$(echo "$matrices" | node -e 'let d="";process.stdin.on("data",c=>d+=c).on("end",()=>console.log(JSON.stringify(JSON.parse(d)[process.argv[1]])))' "$key")" >> "$GITHUB_OUTPUT"
done
linux:
name: ${{ matrix.name }}
# The update/installer route choices map to the linux update methods;
# either way the whole linux matrix runs (legs are cheap and the
# distinction wasn't worth a filter layer in the generator).
if: github.event_name != 'workflow_dispatch' || contains(fromJSON('["all", "both", "update", "installer"]'), inputs.route)
needs: generate-matrix
strategy:
# One leg breaking is worth knowing about even if another already
# failed, so let every leg report.
fail-fast: false
max-parallel: 4
matrix: ${{ fromJSON(needs.generate-matrix.outputs.linux) }}
uses: ./.github/workflows/install-e2e-run.yml
with:
install-method: ${{ matrix.install_method }}
update-method: ${{ matrix.update_method }}
install-ref: ${{ matrix.install_ref }}
windows:
name: ${{ matrix.name }}
if: github.event_name != 'workflow_dispatch' || contains(fromJSON('["all", "windows-desktop"]'), inputs.route)
needs: generate-matrix
strategy:
fail-fast: false
# 2x-cost runners at ~14 min a leg: two at a time is throughput
# enough without hogging the Windows pool.
max-parallel: 2
matrix: ${{ fromJSON(needs.generate-matrix.outputs.windows) }}
uses: ./.github/workflows/install-e2e-windows-run.yml
with:
install-method: ${{ matrix.install_method }}
update-method: ${{ matrix.update_method }}
install-ref: ${{ matrix.install_ref }}
tag-has-desktop: ${{ matrix.tag_has_desktop }}
macos:
name: ${{ matrix.name }}
if: github.event_name != 'workflow_dispatch' || inputs.route == 'all'
needs: generate-matrix
strategy:
fail-fast: false
# 10x-cost runners: keep concurrency low.
max-parallel: 2
matrix: ${{ fromJSON(needs.generate-matrix.outputs.macos) }}
# The same OS-agnostic driver as linux -- only the runner differs.
uses: ./.github/workflows/install-e2e-run.yml
with:
install-method: ${{ matrix.install_method }}
update-method: ${{ matrix.update_method }}
install-ref: ${{ matrix.install_ref }}
runner: macos-latest