Skip to content

fix(terminal): prefer $SHELL over bash for background process spawning - #42219

Closed
liuhao1024 wants to merge 1 commit into
NousResearch:mainfrom
liuhao1024:fix/find-shell-prefer-user-shell
Closed

fix(terminal): prefer $SHELL over bash for background process spawning#42219
liuhao1024 wants to merge 1 commit into
NousResearch:mainfrom
liuhao1024:fix/find-shell-prefer-user-shell

Conversation

@liuhao1024

Copy link
Copy Markdown
Contributor

What does this PR do?

Fixes background process spawning on macOS by preferring the user's $SHELL (typically /bin/zsh) over bash for spawn_local. On macOS Catalina+, _find_shell returned /bin/bash (bash 3.2) which silently swallowed commands when ~/.bash_profile contained exec /bin/zsh -l.

Related Issue

Fixes #42203

Type of Change

  • 🐛 Bug fix (non-breaking change that fixes an issue)

Changes Made

  • tools/environments/local.py: Replace _find_shell = _find_bash alias with a proper function that prefers $SHELL on POSIX, falling back to _find_bash when $SHELL is unset or invalid. On Windows, behavior is unchanged.
  • tests/tools/test_find_shell.py: 7 regression tests covering $SHELL preference, fallback to _find_bash, empty/missing $SHELL, Windows passthrough, and return-type invariant.

How to Test

  1. On macOS, verify _find_shell() returns /bin/zsh (or whatever $SHELL is set to):
    from tools.environments.local import _find_shell
    print(_find_shell())  # Should print /bin/zsh on macOS Catalina+
  2. Unset $SHELL and verify fallback to bash:
    SHELL="" python -c "from tools.environments.local import _find_shell; print(_find_shell())"
  3. Run the new tests:
    pytest tests/tools/test_find_shell.py -v
  4. Run existing process_registry tests to verify no regression:
    pytest tests/tools/test_process_registry.py -v -k "spawn_local"

Checklist

Code

  • I've read the Contributing Guide
  • My commit messages follow Conventional Commits (fix(scope):, feat(scope):, etc.)
  • I searched for existing PRs to make sure this isn't a duplicate
  • My PR contains only changes related to this fix/feature (no unrelated commits)
  • I've run pytest tests/ -q and all tests pass
  • I've added tests for my changes (required for bug fixes, strongly encouraged for features)
  • I've tested on my platform: macOS

Documentation & Housekeeping

  • I've updated relevant documentation (README, docs/, docstrings) — or N/A
  • I've updated cli-config.yaml.example if I added/changed config keys — or N/A
  • I've updated CONTRIBUTING.md or AGENTS.md if I changed architecture or workflows — or N/A
  • I've considered cross-platform impact (Windows, macOS) per the compatibility guide — or N/A
  • I've updated tool descriptions/schemas if I changed tool behavior — or N/A

Code Intelligence

  • Analyzed: _find_shell, _find_bash in tools/environments/local.py; spawn_local in tools/process_registry.py
  • Callers of _find_shell: process_registry.py (PTY spawn at line 550, pipe spawn at line 588)
  • Callers of _find_bash: local.py (_run_bash at line 502), local.py (_find_shell fallback)
  • Blast radius: LOW — _find_shell is only used by spawn_local for background processes; foreground terminal execution uses _find_bash via LocalEnvironment._run_bash
  • Related patterns: The _find_bash function is unchanged — callers that explicitly need bash continue to get it

_find_shell was an alias for _find_bash, which on macOS returns
/bin/bash (bash 3.2) via shutil.which("bash"). When spawn_local uses
this shell with -lic flags and stdin=/dev/null, bash 3.2 sources
~/.bash_profile which on many macOS setups contains "exec /bin/zsh -l".
That exec replaces bash with zsh but drops the -c argument, so the
background command never runs — the subprocess exits 0 with no output.

Change _find_shell to prefer $SHELL on POSIX (typically /bin/zsh on
macOS Catalina+), falling back to _find_bash when $SHELL is unset or
points to a non-existent path. _find_bash is unchanged — callers that
explicitly need bash (like _run_bash) continue to get it.

Fixes NousResearch#42203
@alt-glitch alt-glitch added type/bug Something isn't working P1 High — major feature broken, no workaround comp/tools Tool registry, model_tools, toolsets tool/terminal Terminal execution and process management backend/local Local shell execution labels Jun 8, 2026
@teknium1 teknium1 added sweeper:risk-compatibility Sweeper risk: may break existing users, config, migrations, defaults, or upgrades sweeper:blast-contained Sweeper blast radius: contained — one narrow path / opt-in / few users labels Jun 21, 2026
@kshitijk4poor

Copy link
Copy Markdown
Collaborator

Salvaged into #53106 (rebased onto current main, 506 commits ahead — cherry-pick applied cleanly). All credit to @liuhao1024 for the original fix. Thanks for the contribution!

kshitijk4poor pushed a commit that referenced this pull request Jun 26, 2026
#42203)

On macOS, terminal(background=true) silently failed: the process returned a
session_id and exit_code=0 but the command never ran (empty stdout, no side
effects). Root cause is two interacting issues:

1. _find_shell was aliased to _find_bash, which prefers `shutil.which("bash")`
   → /bin/bash (GNU bash 3.2, still shipped on macOS) over $SHELL (/bin/zsh).
2. process_registry.spawn_local runs [shell, "-lic", "set +m; <cmd>"] with
   stdin=/dev/null. bash 3.2 as a login shell sources ~/.bash_profile, which on
   many macOS setups contains `exec /bin/zsh -l`; that exec replaces bash but
   drops the -c argument, so the command is swallowed (exit 0, no output).

Decouple _find_shell from _find_bash: _find_shell now prefers the user's
configured $SHELL on POSIX (the shell they actually log in with), falling back
to _find_bash when $SHELL is unset/missing. _find_bash is unchanged, so callers
that genuinely need bash (e.g. the _run_bash login-shell snapshot) keep bash
semantics. zsh handles -lic correctly even with redirected stdin.

Salvaged from #42219 by @liuhao1024 (authorship preserved via cherry-pick).
On top of the original (8 unit tests covering $SHELL-set/unset/missing/empty,
Windows-ignores-$SHELL, _find_bash-unchanged), added an E2E regression test
that reproduces the real bash-3.2 login-shell swallow (exit 0 / no file) and
asserts the shell _find_shell selects actually executes a -lic background
command. Mutation-verified: reverting _find_shell to the bash alias fails the
$SHELL-preference test. Bug reproduced directly: /bin/bash 3.2 -lic with a
.bash_profile->exec-zsh creates no file; zsh -lic does.

Closes #42203. Supersedes #42290.
pai-scaffolde pushed a commit to pai-scaffolde/hermes-agent that referenced this pull request Jun 28, 2026
NousResearch#42203)

On macOS, terminal(background=true) silently failed: the process returned a
session_id and exit_code=0 but the command never ran (empty stdout, no side
effects). Root cause is two interacting issues:

1. _find_shell was aliased to _find_bash, which prefers `shutil.which("bash")`
   → /bin/bash (GNU bash 3.2, still shipped on macOS) over $SHELL (/bin/zsh).
2. process_registry.spawn_local runs [shell, "-lic", "set +m; <cmd>"] with
   stdin=/dev/null. bash 3.2 as a login shell sources ~/.bash_profile, which on
   many macOS setups contains `exec /bin/zsh -l`; that exec replaces bash but
   drops the -c argument, so the command is swallowed (exit 0, no output).

Decouple _find_shell from _find_bash: _find_shell now prefers the user's
configured $SHELL on POSIX (the shell they actually log in with), falling back
to _find_bash when $SHELL is unset/missing. _find_bash is unchanged, so callers
that genuinely need bash (e.g. the _run_bash login-shell snapshot) keep bash
semantics. zsh handles -lic correctly even with redirected stdin.

Salvaged from NousResearch#42219 by @liuhao1024 (authorship preserved via cherry-pick).
On top of the original (8 unit tests covering $SHELL-set/unset/missing/empty,
Windows-ignores-$SHELL, _find_bash-unchanged), added an E2E regression test
that reproduces the real bash-3.2 login-shell swallow (exit 0 / no file) and
asserts the shell _find_shell selects actually executes a -lic background
command. Mutation-verified: reverting _find_shell to the bash alias fails the
$SHELL-preference test. Bug reproduced directly: /bin/bash 3.2 -lic with a
.bash_profile->exec-zsh creates no file; zsh -lic does.

Closes NousResearch#42203. Supersedes NousResearch#42290.
waefrebeorn pushed a commit to waefrebeorn/slermes that referenced this pull request Jul 2, 2026
NousResearch#42203)

On macOS, terminal(background=true) silently failed: the process returned a
session_id and exit_code=0 but the command never ran (empty stdout, no side
effects). Root cause is two interacting issues:

1. _find_shell was aliased to _find_bash, which prefers `shutil.which("bash")`
   → /bin/bash (GNU bash 3.2, still shipped on macOS) over $SHELL (/bin/zsh).
2. process_registry.spawn_local runs [shell, "-lic", "set +m; <cmd>"] with
   stdin=/dev/null. bash 3.2 as a login shell sources ~/.bash_profile, which on
   many macOS setups contains `exec /bin/zsh -l`; that exec replaces bash but
   drops the -c argument, so the command is swallowed (exit 0, no output).

Decouple _find_shell from _find_bash: _find_shell now prefers the user's
configured $SHELL on POSIX (the shell they actually log in with), falling back
to _find_bash when $SHELL is unset/missing. _find_bash is unchanged, so callers
that genuinely need bash (e.g. the _run_bash login-shell snapshot) keep bash
semantics. zsh handles -lic correctly even with redirected stdin.

Salvaged from NousResearch#42219 by @liuhao1024 (authorship preserved via cherry-pick).
On top of the original (8 unit tests covering $SHELL-set/unset/missing/empty,
Windows-ignores-$SHELL, _find_bash-unchanged), added an E2E regression test
that reproduces the real bash-3.2 login-shell swallow (exit 0 / no file) and
asserts the shell _find_shell selects actually executes a -lic background
command. Mutation-verified: reverting _find_shell to the bash alias fails the
$SHELL-preference test. Bug reproduced directly: /bin/bash 3.2 -lic with a
.bash_profile->exec-zsh creates no file; zsh -lic does.

Closes NousResearch#42203. Supersedes NousResearch#42290.
habarmc1223-sudo pushed a commit to habarmc1223-sudo/hermes-agent-fluxmem that referenced this pull request Jul 8, 2026
NousResearch#42203)

On macOS, terminal(background=true) silently failed: the process returned a
session_id and exit_code=0 but the command never ran (empty stdout, no side
effects). Root cause is two interacting issues:

1. _find_shell was aliased to _find_bash, which prefers `shutil.which("bash")`
   → /bin/bash (GNU bash 3.2, still shipped on macOS) over $SHELL (/bin/zsh).
2. process_registry.spawn_local runs [shell, "-lic", "set +m; <cmd>"] with
   stdin=/dev/null. bash 3.2 as a login shell sources ~/.bash_profile, which on
   many macOS setups contains `exec /bin/zsh -l`; that exec replaces bash but
   drops the -c argument, so the command is swallowed (exit 0, no output).

Decouple _find_shell from _find_bash: _find_shell now prefers the user's
configured $SHELL on POSIX (the shell they actually log in with), falling back
to _find_bash when $SHELL is unset/missing. _find_bash is unchanged, so callers
that genuinely need bash (e.g. the _run_bash login-shell snapshot) keep bash
semantics. zsh handles -lic correctly even with redirected stdin.

Salvaged from NousResearch#42219 by @liuhao1024 (authorship preserved via cherry-pick).
On top of the original (8 unit tests covering $SHELL-set/unset/missing/empty,
Windows-ignores-$SHELL, _find_bash-unchanged), added an E2E regression test
that reproduces the real bash-3.2 login-shell swallow (exit 0 / no file) and
asserts the shell _find_shell selects actually executes a -lic background
command. Mutation-verified: reverting _find_shell to the bash alias fails the
$SHELL-preference test. Bug reproduced directly: /bin/bash 3.2 -lic with a
.bash_profile->exec-zsh creates no file; zsh -lic does.

Closes NousResearch#42203. Supersedes NousResearch#42290.
santhreal pushed a commit to santhreal/hermes-agent that referenced this pull request Jul 13, 2026
NousResearch#42203)

On macOS, terminal(background=true) silently failed: the process returned a
session_id and exit_code=0 but the command never ran (empty stdout, no side
effects). Root cause is two interacting issues:

1. _find_shell was aliased to _find_bash, which prefers `shutil.which("bash")`
   → /bin/bash (GNU bash 3.2, still shipped on macOS) over $SHELL (/bin/zsh).
2. process_registry.spawn_local runs [shell, "-lic", "set +m; <cmd>"] with
   stdin=/dev/null. bash 3.2 as a login shell sources ~/.bash_profile, which on
   many macOS setups contains `exec /bin/zsh -l`; that exec replaces bash but
   drops the -c argument, so the command is swallowed (exit 0, no output).

Decouple _find_shell from _find_bash: _find_shell now prefers the user's
configured $SHELL on POSIX (the shell they actually log in with), falling back
to _find_bash when $SHELL is unset/missing. _find_bash is unchanged, so callers
that genuinely need bash (e.g. the _run_bash login-shell snapshot) keep bash
semantics. zsh handles -lic correctly even with redirected stdin.

Salvaged from NousResearch#42219 by @liuhao1024 (authorship preserved via cherry-pick).
On top of the original (8 unit tests covering $SHELL-set/unset/missing/empty,
Windows-ignores-$SHELL, _find_bash-unchanged), added an E2E regression test
that reproduces the real bash-3.2 login-shell swallow (exit 0 / no file) and
asserts the shell _find_shell selects actually executes a -lic background
command. Mutation-verified: reverting _find_shell to the bash alias fails the
$SHELL-preference test. Bug reproduced directly: /bin/bash 3.2 -lic with a
.bash_profile->exec-zsh creates no file; zsh -lic does.

Closes NousResearch#42203. Supersedes NousResearch#42290.
Gravezzz pushed a commit to Gravezzz/hermes-agent that referenced this pull request Jul 21, 2026
NousResearch#42203)

On macOS, terminal(background=true) silently failed: the process returned a
session_id and exit_code=0 but the command never ran (empty stdout, no side
effects). Root cause is two interacting issues:

1. _find_shell was aliased to _find_bash, which prefers `shutil.which("bash")`
   → /bin/bash (GNU bash 3.2, still shipped on macOS) over $SHELL (/bin/zsh).
2. process_registry.spawn_local runs [shell, "-lic", "set +m; <cmd>"] with
   stdin=/dev/null. bash 3.2 as a login shell sources ~/.bash_profile, which on
   many macOS setups contains `exec /bin/zsh -l`; that exec replaces bash but
   drops the -c argument, so the command is swallowed (exit 0, no output).

Decouple _find_shell from _find_bash: _find_shell now prefers the user's
configured $SHELL on POSIX (the shell they actually log in with), falling back
to _find_bash when $SHELL is unset/missing. _find_bash is unchanged, so callers
that genuinely need bash (e.g. the _run_bash login-shell snapshot) keep bash
semantics. zsh handles -lic correctly even with redirected stdin.

Salvaged from NousResearch#42219 by @liuhao1024 (authorship preserved via cherry-pick).
On top of the original (8 unit tests covering $SHELL-set/unset/missing/empty,
Windows-ignores-$SHELL, _find_bash-unchanged), added an E2E regression test
that reproduces the real bash-3.2 login-shell swallow (exit 0 / no file) and
asserts the shell _find_shell selects actually executes a -lic background
command. Mutation-verified: reverting _find_shell to the bash alias fails the
$SHELL-preference test. Bug reproduced directly: /bin/bash 3.2 -lic with a
.bash_profile->exec-zsh creates no file; zsh -lic does.

Closes NousResearch#42203. Supersedes NousResearch#42290.
leewenjie pushed a commit to leewenjie/hermes-agent that referenced this pull request Aug 7, 2026
NousResearch#42203)

On macOS, terminal(background=true) silently failed: the process returned a
session_id and exit_code=0 but the command never ran (empty stdout, no side
effects). Root cause is two interacting issues:

1. _find_shell was aliased to _find_bash, which prefers `shutil.which("bash")`
   → /bin/bash (GNU bash 3.2, still shipped on macOS) over $SHELL (/bin/zsh).
2. process_registry.spawn_local runs [shell, "-lic", "set +m; <cmd>"] with
   stdin=/dev/null. bash 3.2 as a login shell sources ~/.bash_profile, which on
   many macOS setups contains `exec /bin/zsh -l`; that exec replaces bash but
   drops the -c argument, so the command is swallowed (exit 0, no output).

Decouple _find_shell from _find_bash: _find_shell now prefers the user's
configured $SHELL on POSIX (the shell they actually log in with), falling back
to _find_bash when $SHELL is unset/missing. _find_bash is unchanged, so callers
that genuinely need bash (e.g. the _run_bash login-shell snapshot) keep bash
semantics. zsh handles -lic correctly even with redirected stdin.

Salvaged from NousResearch#42219 by @liuhao1024 (authorship preserved via cherry-pick).
On top of the original (8 unit tests covering $SHELL-set/unset/missing/empty,
Windows-ignores-$SHELL, _find_bash-unchanged), added an E2E regression test
that reproduces the real bash-3.2 login-shell swallow (exit 0 / no file) and
asserts the shell _find_shell selects actually executes a -lic background
command. Mutation-verified: reverting _find_shell to the bash alias fails the
$SHELL-preference test. Bug reproduced directly: /bin/bash 3.2 -lic with a
.bash_profile->exec-zsh creates no file; zsh -lic does.

Closes NousResearch#42203. Supersedes NousResearch#42290.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

backend/local Local shell execution comp/tools Tool registry, model_tools, toolsets P1 High — major feature broken, no workaround sweeper:blast-contained Sweeper blast radius: contained — one narrow path / opt-in / few users sweeper:risk-compatibility Sweeper risk: may break existing users, config, migrations, defaults, or upgrades tool/terminal Terminal execution and process management type/bug Something isn't working

Projects

None yet

Development

Successfully merging this pull request may close these issues.

macOS: background processes (terminal background=true) silently fail — _find_shell returns bash whose login shell (-l) swallows commands

4 participants