feat(desktop): thin installer + first-launch install.ps1 bootstrap

Converges the Windows packaged desktop installer onto a single canonical
install topology: drop the Electron shell only (~80MB instead of ~500MB),
clone Hermes Agent at a build-time-pinned commit on first launch via
install.ps1's stage protocol, and treat the resulting git checkout at
%LOCALAPPDATA%\hermes\hermes-agent\ as the canonical install location
(same path the CLI installer uses).  Future updates flow through the
existing applyUpdates() git-pull path.

Replaces the previous fat-installer architecture where the .exe bundled
a pre-staged hermes-agent source tree under resources/hermes-agent/ that
was then sync'd into ACTIVE_HERMES_ROOT at launch -- a complicated
factory-vs-active dance with several footguns (FACTORY_HERMES_ROOT
mismatch on path resolve, isGitCheckout guard regressions, pyproject
hash drift detection inside the sync loop).

Architecture overview
---------------------

  Build time
    apps/desktop/scripts/write-build-stamp.cjs writes
    apps/desktop/build/install-stamp.json with {commit, branch, builtAt,
    dirty}.  Honours $GITHUB_SHA / $GITHUB_REF_NAME in CI, falls back to
    `git rev-parse HEAD` locally.

    apps/desktop/scripts/stage-native-deps.cjs copies the runtime subset
    of @homebridge/node-pty-prebuilt-multiarch from the workspace-root
    node_modules into apps/desktop/build/native-deps/.  Workspace dedup
    hoists this dep to the root, out of reach of electron-builder's
    `files:`-restricted collector; staging gives us a deterministic
    path to extraResources.

    electron-builder ships both into resources/install-stamp.json and
    resources/native-deps/ respectively.

  Boot resolver (electron/main.cjs)
    Resolver order:
      1. HERMES_DESKTOP_HERMES_ROOT override
      2. SOURCE_REPO_ROOT (dev mode)
      3. ACTIVE_HERMES_ROOT git checkout WITH .hermes-bootstrap-complete
         marker -- the post-install fast path
      4. `hermes` on PATH (CLI-installed user adding the desktop)
      5. pip-installed hermes_cli via system Python
      6. bootstrap-needed sentinel -> hand off to runBootstrap

    Deletes the entire FACTORY_HERMES_ROOT / RUNTIME_MARKER /
    syncTreeExcludingVenv machinery (-200 lines).  The isGitCheckout
    guard that bit us in the install.ps1 PR is gone.

  First-launch bootstrap (electron/bootstrap-runner.cjs)
    1. Resolve install.ps1: prefer SOURCE_REPO_ROOT/scripts (dev), else
       download from GitHub raw at INSTALL_STAMP.commit (cached at
       HERMES_HOME\bootstrap-cache\install-<sha>.ps1).
    2. Fetch the stage manifest via install.ps1 -Manifest -Commit X
       -Branch Y.
    3. Iterate stages: install.ps1 -Stage <name> -NonInteractive -Json
       -Commit X -Branch Y per stage.
    4. On all stages green: write the .hermes-bootstrap-complete
       marker with {schemaVersion, pinnedCommit, pinnedBranch,
       completedAt, desktopVersion}.

    Per-run log to HERMES_HOME\logs\bootstrap-<ts>.log.  Cancellation
    via AbortSignal.  Manifest cache so retries don't re-download.

  Install overlay (src/components/desktop-install-overlay.tsx)
    Mounted alongside the existing onboarding overlay; flexbox card
    with header (static) + middle (scrollable) + footer (failure-only,
    static).  Subscribes to hermes:bootstrap:event IPC + resyncs from
    hermes:bootstrap:get on mount/reload.  Renders:
      - 14-stage checklist with per-stage state icons
      - Overall progress bar + current-stage spotlight
      - Auto-expanded installer-output panel on failure
      - "Copy output" button (full ring buffer + error to clipboard)
      - "Reload and retry" wired through hermes:bootstrap:reset to
        clear main.cjs's latched failure
    Synthetic empty-manifest event from main.cjs flips the overlay to
    'active' immediately so the slow install.ps1 download doesn't
    leave the user staring at the generic Preparing splash.

  Failure latching (main.cjs)
    bootstrapFailure module-scope variable holds the rejection after
    install.ps1 fails.  startHermes() throws the latched error
    immediately when set, bypassing the entire ensureRuntime +
    runBootstrap chain.  Without this, the renderer's ensureGatewayOpen
    retries would re-run install.ps1 in a 5-10 min hot loop while the
    user was still reading the failure overlay.  Cleared via
    hermes:bootstrap:reset on user-driven retry.

  Unsupported-platform overlay (1F)
    macOS / Linux packaged builds (no install.sh stage protocol yet)
    emit an unsupported-platform event with a copy-pasteable install
    command + docs URL.  Dedicated overlay branch with "Copy command"
    + "I've run it -- retry" buttons.

install.ps1 additions (Phase 1F.3 + 1F.5)
-----------------------------------------

  New -Commit and -Tag string params.  Precedence Commit > Tag >
  Branch.  Honoured by all three code paths (update / fresh clone /
  ZIP fallback), with archive URL selection that handles each
  ref-type variant.  Detached-HEAD checkouts intentionally -- they're
  pins, not branches the user pulls into.

  EAP=Continue wrap around the new pin-step git invocations.  `git
  fetch origin <commit>` writes the routine 'From <url>' info line to
  stderr; under the script's global EAP=Stop that terminates the
  script even though fetch+checkout succeed.  Matches the established
  pattern in Install-Uv, Test-Python, _Run-NpmInstall.

Backend fix (hermes_cli/web_server.py)
--------------------------------------

  CORS allow_origin_regex now accepts Origin: 'null'.  Packaged
  Electron loads index.html via file://; Chromium sets the WebSocket
  upgrade Origin header to the opaque origin 'null', which the old
  regex rejected with HTTP 403 before gateway_ws() ever ran.  This
  failure mode was masked in the older FACTORY_HERMES_ROOT
  architecture because the resolver often found an existing hermes
  on PATH with different binding behavior.

  Security maintained: localhost-only bind keeps cross-machine pages
  out; per-process session token still gates every authenticated
  /api/ endpoint regardless of Origin.

Desktop QoL
-----------

  DevTools is now enabled in packaged builds (F12 / Cmd+Opt+I).
  Field-debugging trade-off: tiny attack surface increase versus
  a much better support story when CSP / WS / theme issues surface.

  NSIS prereq-check page deleted (-767 lines).  The standard
  Welcome -> License -> Directory -> InstallFiles -> Finish wizard
  now installs without custom Python/Git/ripgrep detection -- those
  prereqs are install.ps1's job at first launch.

Test infrastructure (Phase 1G)
------------------------------

  apps/desktop/scripts/test-desktop.mjs rewritten as a cross-platform
  bundle validator (was darwin-only and asserted on dead factory-
  payload paths):
    NEGATIVE: hermes_cli/main.py is NOT shipped (regression guard)
    POSITIVE: install-stamp.json carries a real commit + branch
    POSITIVE: node-pty native deps shipped under resources/native-deps
    POSITIVE: renderer dist/index.html reachable (asar or unpacked)
  New nsis mode and npm run test:desktop:nsis script.

Validated end-to-end on clean Win10 VM
--------------------------------------

  Confirmed: NSIS installer drops Electron shell, app launches,
  install overlay shows progress, install.ps1 clones the pinned
  commit, 14 stages run to completion, marker written, backend
  spawns, WebSocket connects, onboarding overlay asks for API key,
  main UI loads, integrated terminal works.

  Failures handled: bootstrap stays failed (no hot-loop retry),
  "Copy output" gives actionable transcript, "Reload and retry"
  explicitly re-runs install.ps1.

What's deferred
---------------

  - MSIX wrapping (Phase 2): same Electron .exe under MSIX manifest
    with runFullTrust, signed and submitted to Microsoft Store.
  - install.sh stage protocol parity (Phase 2): once shipped, the
    unsupported-platform overlay becomes drive-it-yourself and
    macOS/Linux packaged installers gain feature parity with Windows.
This commit is contained in:
emozilla
2026-05-18 02:26:46 -04:00
parent 046f0c01cb
commit 705eaa054a
17 changed files with 2072 additions and 1212 deletions
+23 -21
View File
@@ -97,7 +97,7 @@ npm run dist:mac:zip # zip only
npm run dist:win # NSIS + MSI
```
Before packaging, `stage:hermes` copies the Python Hermes payload into `build/hermes-agent`. Electron Builder then ships it as `Contents/Resources/hermes-agent`.
Before packaging, the desktop app no longer bundles a copy of the Hermes Agent Python source. Instead, the packaged Electron app will fetch and install Hermes Agent at first launch via `scripts/install.ps1`'s stage protocol (Windows) — see the bootstrap flow documented in `electron/main.cjs`. macOS and Linux packaged builds are temporarily non-functional until `install.sh` gains the same stage protocol; dev workflows on all three platforms continue to work since they resolve a sibling source checkout.
## Automated Releases
@@ -187,12 +187,14 @@ Hermes Desktop shares its install layout with the CLI installers (`scripts/insta
```text
HERMES_HOME/ # %LOCALAPPDATA%\hermes (Windows)
# ~/.hermes (macOS / Linux)
├── hermes-agent/ # ACTIVE_HERMES_ROOT — the canonical install
├── hermes-agent/ # ACTIVE_HERMES_ROOT — git checkout
│ ├── .git/ # canonical install is always a git checkout
│ ├── hermes_cli/, agent/, ... # Python source
│ ├── pyproject.toml # source of truth for deps
│ ├── venv/ # virtualenv (Scripts\python.exe on Windows,
│ │ # bin/python elsewhere)
│ └── .hermes-desktop-runtime.json # marker: schema version + pyproject hash
│ └── .hermes-bootstrap-complete # marker: first-launch install.ps1 succeeded
├── git/ # PortableGit (Windows; installed by install.ps1)
├── config.yaml # user config
├── .env # API keys
└── logs/
@@ -202,33 +204,33 @@ HERMES_HOME/ # %LOCALAPPDATA%\hermes (Windows)
└── gateway.log
```
The factory image (`Contents/Resources/hermes-agent` on macOS, `resources\hermes-agent` on Windows) ships inside the `.app` / `.exe` and seeds `HERMES_HOME/hermes-agent` on first launch.
The packaged installer ships only the Electron app — Hermes Agent itself is fetched and installed at first launch by running `scripts/install.ps1` (Windows) against the git ref baked into the .exe at build time (see `apps/desktop/scripts/write-build-stamp.cjs`).
### Resolution order
The desktop resolves a Hermes backend in this order:
1. `HERMES_DESKTOP_HERMES_ROOT` — explicit dev override.
2. Existing `hermes` CLI on PATH (skipped when `HERMES_DESKTOP_IGNORE_EXISTING=1`).
3. Repo source root — only when running `npm run dev` from a checkout. Takes precedence over `HERMES_HOME/hermes-agent` so devs always run their local edits.
4. `HERMES_HOME/hermes-agent` if it already exists (CLI installer or prior desktop launch).
5. Packaged + factory image present → sync factory → `HERMES_HOME/hermes-agent`, then use it.
6. Pip-installed `hermes_cli` module via system Python.
2. Repo source root — only when running `npm run dev` from a checkout. Takes precedence over `HERMES_HOME/hermes-agent` so devs always run their local edits.
3. `HERMES_HOME/hermes-agent` if the `.hermes-bootstrap-complete` marker is present. The marker attests that install.ps1 succeeded and the user finished initial configuration; we trust the install and skip the bootstrap flow on every launch after the first.
4. Existing `hermes` CLI on PATH (skipped when `HERMES_DESKTOP_IGNORE_EXISTING=1`).
5. Pip-installed `hermes_cli` module via system Python.
6. None of the above → bootstrap-needed sentinel. The desktop's first-launch wizard runs `scripts/install.ps1` stages, then writes the marker on success.
### First-launch flow on a packaged install
1. Sync factory image → `HERMES_HOME/hermes-agent`. Skipped if a `.git` directory exists at the destination (developer install) — never overwrites a user's local repo.
2. Create venv at `HERMES_HOME/hermes-agent/venv` using system Python (errors out with a Python-install hint if no Python 3.11+ is found).
3. `pip install -e HERMES_HOME/hermes-agent``pyproject.toml` is the single source of truth for dependencies.
4. Stamp `.hermes-desktop-runtime.json` with the schema version + pyproject hash + factory version.
1. `resolveHermesBackend()` returns `kind: 'bootstrap-needed'`.
2. The renderer shows the install overlay; main fetches `scripts/install.ps1` from GitHub at the pinned commit (from `install-stamp.json`).
3. Main drives `install.ps1 -Manifest` to get the stage list, then iterates `install.ps1 -Stage <name> -NonInteractive -Json` with live progress events to the renderer.
4. On all stages succeeding, main writes `.hermes-bootstrap-complete` with `{ schemaVersion, pinnedCommit, pinnedBranch, completedAt, desktopVersion }`.
5. Renderer hands off to the existing onboarding overlay (API key / model / persona).
6. Subsequent launches see the marker and skip everything in steps 1-5.
Subsequent launches compare the marker against the active `pyproject.toml` and skip steps 2-4 when nothing has changed.
### Updates
### Upgrades
Once bootstrapped, the install is a real git checkout. Updates flow through the in-app update path (`applyUpdates()``git fetch && git pull --ff-only` against the configured branch) or `hermes update` from the CLI. Both check `pyproject.toml` drift and re-run `pip install -e .` only when needed.
A new installer drops a new factory image. On next launch the marker mismatches → factory contents are copied over `HERMES_HOME/hermes-agent` (excluding `venv/`, `.git`, `__pycache__`, etc.), `pip install -e` re-runs to pick up new deps, the marker is re-stamped. The venv is preserved across upgrades to keep the upgrade fast when deps haven't moved.
A user who installed via `scripts/install.ps1` / `scripts/install.sh` (so `HERMES_HOME/hermes-agent/.git` exists) is detected as a developer install and the desktop never overwrites their checkout — they keep using `hermes update` / `git pull` to update.
A user who installed via `scripts/install.ps1` directly (so `HERMES_HOME/hermes-agent/.git` exists but no `.hermes-bootstrap-complete` marker) is detected via resolver step 4 (their `hermes` CLI on PATH) and the desktop reuses their install without re-running the bootstrap.
## Debugging
@@ -241,14 +243,14 @@ HERMES_HOME/logs/desktop.log # %LOCALAPPDATA%\hermes\logs\desktop.log on Win
If the UI reports `Desktop boot failed`, check that log first. It includes the backend command output and recent Python traceback context.
To reset desktop runtime state (forces re-sync from the factory image and re-`pip install -e .` on next launch):
To force a fresh first-launch bootstrap (rare — useful for development / dogfooding the install flow):
```bash
# macOS / Linux
rm "$HOME/.hermes/hermes-agent/.hermes-desktop-runtime.json"
rm "$HOME/.hermes/hermes-agent/.hermes-bootstrap-complete"
# Windows (PowerShell)
Remove-Item "$env:LOCALAPPDATA\hermes\hermes-agent\.hermes-desktop-runtime.json"
Remove-Item "$env:LOCALAPPDATA\hermes\hermes-agent\.hermes-bootstrap-complete"
```
For a full reset of just the Python venv (rare — usually only needed if the venv is broken):