fix(tools): install cua-driver when Computer Use is enabled via 'hermes tools' (#22765)

Returning users who enabled '🖱️ Computer Use (macOS)' via 'hermes tools'
saw '✓ Saved configuration' but no install — cua-driver was never on
PATH and the toolset failed at first use. Two compounding causes:

1. _toolset_needs_configuration_prompt fell through to _toolset_has_keys,
   which returned True for any provider with empty env_vars. cua-driver
   has no env vars, so the gate skipped _configure_toolset entirely and
   _run_post_setup('cua_driver') never ran.

2. No stable CLI entry-point existed for re-running the install when
   the picker no-op'd it (e.g. when toggling the toolset off+on inside
   one picker session, where 'added' is empty).

Changes:

- hermes_cli/tools_config.py: add _POST_SETUP_INSTALLED registry
  mapping post_setup keys to installed-state predicates. The gate
  now returns True when any visible provider has a registered
  post_setup whose predicate fails. cua_driver is the only opt-in
  for now; other post_setup hooks keep their existing behaviour.
- hermes_cli/main.py: add 'hermes computer-use install' and
  'hermes computer-use status' as a stable docs target. install
  reuses the same _run_post_setup('cua_driver') path that the
  picker invokes; status reports whether cua-driver is on PATH.
- tools/computer_use/cua_backend.py: install hint now points users
  at 'hermes computer-use install' first.
- website/docs/user-guide/features/computer-use.md: document the
  new command as the primary install path.
- website/docs/reference/cli-commands.md: catalog 'hermes
  computer-use' alongside 'hermes tools'.
- tests/hermes_cli/test_post_setup_gating.py: regression coverage
  for the gate predicate (missing -> setup forced, installed ->
  setup skipped, broken predicate -> non-blocking, unregistered
  keys -> behaviour unchanged).

Fixes #22737. Reported by @f-trycua.
This commit is contained in:
Teknium 2026-05-09 13:02:25 -07:00 committed by GitHub
parent 6e5489c9f3
commit 8f711f79a4
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
6 changed files with 205 additions and 4 deletions

View file

@ -8886,6 +8886,7 @@ def _build_provider_choices() -> list[str]:
_BUILTIN_SUBCOMMANDS = frozenset(
{
"acp", "auth", "backup", "checkpoints", "claw", "completion",
"computer-use",
"config", "cron", "curator", "dashboard", "debug", "doctor",
"dump", "fallback", "gateway", "hooks", "import", "insights",
"kanban", "login", "logout", "logs", "mcp", "memory", "model",
@ -10506,6 +10507,54 @@ Examples:
tools_command(args)
tools_parser.set_defaults(func=cmd_tools)
# =========================================================================
# computer-use command — manage Computer Use (cua-driver) on macOS
# =========================================================================
computer_use_parser = subparsers.add_parser(
"computer-use",
help="Manage the Computer Use (cua-driver) backend (macOS)",
description=(
"Install or check the cua-driver binary used by the\n"
"`computer_use` toolset. macOS-only.\n\n"
"Use `hermes computer-use install` to fetch and run the\n"
"upstream cua-driver installer. This is equivalent to the\n"
"post-setup hook that `hermes tools` runs when you first\n"
"enable the Computer Use toolset, and is a stable target\n"
"for re-running the install if it didn't fire (e.g. when\n"
"toggling the toolset on a returning-user setup)."
),
)
computer_use_sub = computer_use_parser.add_subparsers(dest="computer_use_action")
computer_use_sub.add_parser(
"install",
help="Install or repair the cua-driver binary (macOS)",
)
computer_use_sub.add_parser(
"status",
help="Print whether cua-driver is installed and on PATH",
)
def cmd_computer_use(args):
action = getattr(args, "computer_use_action", None)
if action == "install":
from hermes_cli.tools_config import _run_post_setup
_run_post_setup("cua_driver")
return
if action == "status":
import shutil
path = shutil.which("cua-driver")
if path:
print(f"cua-driver: installed at {path}")
return
print("cua-driver: not installed")
print(" Run: hermes computer-use install")
return
# No subcommand → show help
computer_use_parser.print_help()
computer_use_parser.set_defaults(func=cmd_computer_use)
# =========================================================================
# mcp command — manage MCP server connections
# =========================================================================

View file

@ -12,6 +12,7 @@ the `platform_toolsets` key.
import json as _json
import logging
import os
import shutil
import sys
from pathlib import Path
from typing import Dict, List, Optional, Set
@ -1424,12 +1425,52 @@ def _visible_providers(cat: dict, config: dict) -> list[dict]:
return visible
_POST_SETUP_INSTALLED: dict = {
# post_setup_key -> predicate(): True when the install side-effect
# is already satisfied. Used by `_toolset_needs_configuration_prompt`
# to force the provider-setup flow when a no-key provider still needs
# a binary/dependency install (otherwise an already-configured user
# who toggles the toolset on via `hermes tools` gets a silent no-op
# because the gate sees "no env vars to ask about" and skips the
# provider-setup flow that would have run the post_setup hook).
#
# Only entries here are gated; other post_setup hooks (kittentts,
# piper, agent_browser, etc.) keep their existing behaviour. Add an
# entry when (a) the post_setup is the ONLY install side-effect for
# a no-key provider, and (b) an installed-state check is cheap and
# doesn't trigger a heavy import.
"cua_driver": lambda: bool(shutil.which("cua-driver")),
}
def _post_setup_already_installed(post_setup_key: str) -> bool:
"""Return True when the post_setup install side-effect is satisfied."""
predicate = _POST_SETUP_INSTALLED.get(post_setup_key)
if predicate is None:
# No install-state check registered → assume satisfied (don't
# change behaviour for hooks we haven't explicitly opted in).
return True
try:
return bool(predicate())
except Exception:
return True
def _toolset_needs_configuration_prompt(ts_key: str, config: dict) -> bool:
"""Return True when enabling this toolset should open provider setup."""
cat = TOOL_CATEGORIES.get(ts_key)
if not cat:
return not _toolset_has_keys(ts_key, config)
# If any visible provider has a registered post_setup install-state
# check that hasn't been satisfied (e.g. cua-driver binary not on
# PATH yet), force the configuration flow so `_configure_provider`
# invokes `_run_post_setup` and the install actually runs.
for provider in _visible_providers(cat, config):
post_setup = provider.get("post_setup")
if post_setup and not _post_setup_already_installed(post_setup):
return True
if ts_key == "tts":
tts_cfg = config.get("tts", {})
return not isinstance(tts_cfg, dict) or "provider" not in tts_cfg

View file

@ -0,0 +1,71 @@
"""Tests for the post_setup install-state gate in `_toolset_needs_configuration_prompt`.
Regression coverage for the cua-driver silent-no-op bug (issue #22737).
When a no-key provider's only install side-effect is a `post_setup` hook
(cua-driver, etc.), the gate function used to fall through to the
`_toolset_has_keys` catch-all, which returned True for any provider with
empty `env_vars` causing `hermes tools` to write the toolset to config
and exit ` Saved` without ever invoking the post_setup install. These
tests pin the new predicate-aware behaviour so the regression doesn't
sneak back in.
"""
from __future__ import annotations
class TestPostSetupGate:
def test_cua_driver_missing_forces_setup(self, monkeypatch, tmp_path):
"""When cua-driver isn't on PATH, the gate must return True so the
provider-setup flow runs and triggers `_run_post_setup`."""
from hermes_cli import tools_config
monkeypatch.setenv("HERMES_HOME", str(tmp_path))
monkeypatch.setattr(tools_config.shutil, "which", lambda name: None)
assert tools_config._toolset_needs_configuration_prompt(
"computer_use", {}
) is True
def test_cua_driver_installed_skips_setup(self, monkeypatch, tmp_path):
"""When cua-driver is already on PATH, the gate must return False
so a re-save through `hermes tools` doesn't re-prompt the user."""
from hermes_cli import tools_config
monkeypatch.setenv("HERMES_HOME", str(tmp_path))
monkeypatch.setattr(
tools_config.shutil,
"which",
lambda name: "/usr/local/bin/cua-driver" if name == "cua-driver" else None,
)
assert tools_config._toolset_needs_configuration_prompt(
"computer_use", {}
) is False
def test_post_setup_predicate_exception_does_not_block(self, monkeypatch):
"""A predicate that raises must be treated as 'satisfied' so a
broken check can't strand the user in an infinite setup loop."""
from hermes_cli import tools_config
def _boom():
raise RuntimeError("predicate broken")
monkeypatch.setitem(tools_config._POST_SETUP_INSTALLED, "cua_driver", _boom)
assert tools_config._post_setup_already_installed("cua_driver") is True
def test_unregistered_post_setup_treated_as_satisfied(self):
"""post_setup keys without a registered predicate must default to
'satisfied' so we don't change behaviour for hooks we haven't
explicitly opted in (kittentts, piper, agent_browser, etc.)."""
from hermes_cli import tools_config
assert tools_config._post_setup_already_installed("does_not_exist") is True
def test_cua_driver_predicate_registered(self):
"""Keep an explicit pin on the cua_driver entry so accidental
deletion of the registry row would fail this test rather than
silently restore the original silent-no-op bug."""
from hermes_cli import tools_config
assert "cua_driver" in tools_config._POST_SETUP_INSTALLED

View file

@ -84,7 +84,9 @@ def cua_driver_binary_available() -> bool:
def cua_driver_install_hint() -> str:
return (
"cua-driver is not installed. Install with:\n"
"cua-driver is not installed. Install with one of:\n"
" hermes computer-use install\n"
"Or run the upstream installer directly:\n"
' /bin/bash -c "$(curl -fsSL '
'https://raw.githubusercontent.com/trycua/cua/main/libs/cua-driver/scripts/install.sh)"\n'
"Or run `hermes tools` and enable the Computer Use toolset to install it automatically."

View file

@ -66,6 +66,7 @@ hermes [global-options] <command> [subcommand/options]
| `hermes mcp` | Manage MCP server configurations and run Hermes as an MCP server. |
| `hermes plugins` | Manage Hermes Agent plugins (install, enable, disable, remove). |
| `hermes tools` | Configure enabled tools per platform. |
| `hermes computer-use` | Install or check the cua-driver backend (macOS Computer Use). |
| `hermes sessions` | Browse, export, prune, rename, and delete sessions. |
| `hermes insights` | Show token/cost/activity analytics. |
| `hermes fallback` | Interactive manager for the fallback provider chain. |
@ -958,6 +959,26 @@ hermes tools [--summary]
Without `--summary`, this launches the interactive per-platform tool configuration UI.
## `hermes computer-use`
```bash
hermes computer-use <subcommand>
```
Subcommands:
| Subcommand | Description |
|------------|-------------|
| `install` | Run the upstream cua-driver installer (macOS only). |
| `status` | Print whether `cua-driver` is on `$PATH`. |
`hermes computer-use install` is the stable entry point for installing the
[cua-driver](https://github.com/trycua/cua) binary used by the
`computer_use` toolset. It runs the same upstream installer that
`hermes tools` invokes when you first enable Computer Use, so it's safe
to use for re-running the install if the toolset toggle didn't trigger
it (for example, on returning-user setups).
## `hermes sessions`
```bash

View file

@ -27,9 +27,25 @@ cua-driver is the open-source equivalent.
## Enabling
Pick whichever path is most convenient — both run the same upstream installer:
**Option 1: dedicated CLI command (most direct).**
```
hermes computer-use install
```
This fetches and runs the upstream cua-driver installer:
`curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/libs/cua-driver/scripts/install.sh`.
Use `hermes computer-use status` to verify the install.
**Option 2: enable the toolset interactively.**
1. Run `hermes tools`, pick `🖱️ Computer Use (macOS)``cua-driver (background)`.
2. The setup runs the upstream installer:
`curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/libs/cua-driver/scripts/install.sh`.
2. The setup runs the upstream installer (same as Option 1).
After installing, regardless of which path you took:
3. Grant macOS permissions when prompted:
- **System Settings → Privacy & Security → Accessibility** → allow the
terminal (or Hermes app).
@ -143,7 +159,8 @@ HERMES_COMPUTER_USE_BACKEND=noop # records calls, no side effects
## Troubleshooting
**`computer_use backend unavailable: cua-driver is not installed`** — Run
`hermes tools` and enable Computer Use.
`hermes computer-use install` to fetch the cua-driver binary, or run
`hermes tools` and enable the Computer Use toolset.
**Clicks seem to have no effect** — Capture and verify. A modal you
didn't see may be blocking input. Dismiss it with `escape` or the close