- A failed bootout/bootstrap should surface as an error or a false return value. - The install/repair command should not claim success when the reload failed.

hermes - ✅(Solved) Fix [Bug]: launchd plist refresh reports success even when launchctl reload fails [1 pull requests, 1 comments, 2 participants]

hermes2026-04-20 06:36:05

ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

GitHub issue URL

Helpful · Quick feedback

GitHub stats

NousResearch/hermes-agent#12866•Fetched 2026-04-20 12:16:33

View on GitHub

Comments

Participants

Timeline

Reactions

Author

NewTurn2017

Participants

NewTurn2017

ygd58

Timeline (top)

commented ×1cross-referenced ×1referenced ×1

Refreshing an outdated launchd plist always reports success even if launchctl bootout or launchctl bootstrap fails. That makes hermes gateway install/repair silently claim the service definition was updated while launchd may still be unloaded or broken.

Error Message

A failed bootout/bootstrap should surface as an error or a false return value.

Root Cause

Fix Action

Fix / Workaround

Minimal reproduction

from pathlib import Path
from tempfile import TemporaryDirectory
from types import SimpleNamespace
from unittest.mock import patch
from hermes_cli import gateway as g

with patch.object(g, "get_launchd_plist_path", return_value=plist), \
         patch.object(g, "launchd_plist_is_current", return_value=False), \
         patch.object(g, "generate_launchd_plist", return_value="new"), \
         patch.object(g, "get_launchd_label", return_value="com.hermes.agent"), \
         patch.object(g, "_launchd_domain", return_value="gui/501"), \
         patch("subprocess.run", side_effect=fake_run), \
         patch("builtins.print", side_effect=lambda *a, **k: printed.append(" ".join(str(x) for x in a))):
        result = g.refresh_launchd_plist_if_needed()

PR fix notes

PR #12882: fix(gateway): surface launchctl bootstrap failures in refresh_launchd_plist_if_needed

Repository: NousResearch/hermes-agent
Author: ygd58
State: open | merged: False
Link: https://github.com/NousResearch/hermes-agent/pull/12882

Description (problem / solution / changelog)

Problem

refresh_launchd_plist_if_needed() ran both launchctl commands with check=False, ignored their return codes, and always printed a success message and returned True — even when launchctl bootstrap failed.

This means hermes gateway install / repair silently claims the service definition was updated while launchd may still be unloaded or broken.

Fix

bootout: keep check=False — non-zero is harmless when the job was never loaded
bootstrap: capture returncode; return False and print a diagnostic message when non-zero, so install/repair does not claim success when launchd rejected the plist

Tests

Adds tests/hermes_cli/test_launchd_plist_refresh.py with 6 regression tests — all pass:

bootstrap success → returns True
bootstrap failure (rc=5) → returns False, no success message printed
bootout failure → tolerated, still returns True
plist already current → returns False
plist missing → returns False
bootstrap called with check=False

Fixes #12866

Changed files

hermes_cli/gateway.py (modified, +26/-3)
tests/hermes_cli/test_launchd_plist_refresh.py (added, +116/-0)

Code Example

from pathlib import Path
from tempfile import TemporaryDirectory
from types import SimpleNamespace
from unittest.mock import patch
from hermes_cli import gateway as g

calls = []
printed = []
with TemporaryDirectory() as td:
    plist = Path(td) / "com.hermes.agent.plist"
    plist.write_text("old", encoding="utf-8")

    def fake_run(cmd, check=False, timeout=None):
        calls.append({"cmd": cmd, "check": check, "timeout": timeout})
        return SimpleNamespace(returncode=5)

    with patch.object(g, "get_launchd_plist_path", return_value=plist), \
         patch.object(g, "launchd_plist_is_current", return_value=False), \
         patch.object(g, "generate_launchd_plist", return_value="new"), \
         patch.object(g, "get_launchd_label", return_value="com.hermes.agent"), \
         patch.object(g, "_launchd_domain", return_value="gui/501"), \
         patch("subprocess.run", side_effect=fake_run), \
         patch("builtins.print", side_effect=lambda *a, **k: printed.append(" ".join(str(x) for x in a))):
        result = g.refresh_launchd_plist_if_needed()

print(result)   # actual: True
print(printed)  # actual: success message still emitted

RAW_BUFFERClick to expand / collapse

Summary

Affected code

hermes_cli/gateway.py:1334-1340
hermes_cli/gateway.py:1346-1351
coverage gap: tests/hermes_cli/test_gateway_service.py:194-216

Why this is a bug

refresh_launchd_plist_if_needed() rewrites the plist and then runs both launchctl commands with check=False, ignoring non-zero return codes. The function always prints a success message and returns True.

Minimal reproduction

from pathlib import Path
from tempfile import TemporaryDirectory
from types import SimpleNamespace
from unittest.mock import patch
from hermes_cli import gateway as g

calls = []
printed = []
with TemporaryDirectory() as td:
    plist = Path(td) / "com.hermes.agent.plist"
    plist.write_text("old", encoding="utf-8")

    def fake_run(cmd, check=False, timeout=None):
        calls.append({"cmd": cmd, "check": check, "timeout": timeout})
        return SimpleNamespace(returncode=5)

    with patch.object(g, "get_launchd_plist_path", return_value=plist), \
         patch.object(g, "launchd_plist_is_current", return_value=False), \
         patch.object(g, "generate_launchd_plist", return_value="new"), \
         patch.object(g, "get_launchd_label", return_value="com.hermes.agent"), \
         patch.object(g, "_launchd_domain", return_value="gui/501"), \
         patch("subprocess.run", side_effect=fake_run), \
         patch("builtins.print", side_effect=lambda *a, **k: printed.append(" ".join(str(x) for x in a))):
        result = g.refresh_launchd_plist_if_needed()

print(result)   # actual: True
print(printed)  # actual: success message still emitted

Expected behavior

A failed bootout/bootstrap should surface as an error or a false return value.
The install/repair command should not claim success when the reload failed.

Actual behavior

Reload failures are swallowed and reported as success.

Suggested investigation

Check launchctl return codes (or use check=True with targeted handling for harmless bootout cases).
Add a regression test covering non-zero bootstrap return codes.

extent analysis

TL;DR

The refresh_launchd_plist_if_needed function should be modified to handle non-zero return codes from launchctl commands to accurately report failures.

Guidance

Modify the refresh_launchd_plist_if_needed function to set check=True when running launchctl commands to raise an exception on non-zero return codes.
Add targeted exception handling for harmless bootout cases to prevent false positives.
Implement a regression test in tests/hermes_cli/test_gateway_service.py to cover non-zero bootstrap return codes.
Review the hermes_cli/gateway.py code to ensure that error handling is properly propagated to the install and repair commands.

Example

try:
    subprocess.run(["launchctl", "bootout", "..."], check=True)
except subprocess.CalledProcessError as e:
    # Handle harmless bootout cases or re-raise the exception
    if e.returncode == 5:  # Example harmless return code
        print("Harmless bootout error")
    else:
        raise

Notes

The suggested changes assume that the subprocess module is used to run launchctl commands. The actual implementation may vary depending on the specific code and requirements.

Recommendation

Apply workaround: Modify the refresh_launchd_plist_if_needed function to handle non-zero return codes and add regression testing to ensure the fix is effective. This approach addresses the root cause of the issue and provides a clear path to resolving the problem.

Vote matrix · Quick signals

Works

Did the solution work? Tap to confirm.

Easy Fix

Was it a quick fix?

Time Saver

Did it save you time?

Blocking

Was it severely blocking?

Common Issue

Are others likely hitting this too?

Flaky / Intermittent

Is it intermittent?

Verified / Reproducible

Can you reproduce it reliably?

FAQ

Expected behavior

A failed bootout/bootstrap should surface as an error or a false return value.
The install/repair command should not claim success when the reload failed.

#logging issue #authentication issue #prompt issue #agent setup #task chaining

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

hermes - ✅(Solved) Fix [Bug]: launchd plist refresh reports success even when launchctl reload fails [1 pull requests, 1 comments, 2 participants]

Recommended Tools

GitHub issue graph ai analysis

Error Message

Root Cause

Fix Action

Fix / Workaround

Minimal reproduction

PR fix notes

PR #12882: fix(gateway): surface launchctl bootstrap failures in refresh_launchd_plist_if_needed

Description (problem / solution / changelog)

Problem

Fix

Tests

Changed files

Code Example

Summary

Affected code

Why this is a bug

Minimal reproduction

Expected behavior

Actual behavior

Suggested investigation

extent analysis

TL;DR

Guidance

Example

Notes

Recommendation

FAQ

Expected behavior

Still need to ship something?

RELATED_DISCOVERY

TRENDING