fix(mcp): prevent OAuth server from blocking Hermes startup#4465
Closed
heathley wants to merge 1 commit into
Closed
fix(mcp): prevent OAuth server from blocking Hermes startup#4465heathley wants to merge 1 commit into
heathley wants to merge 1 commit into
Conversation
If an MCP server requires OAuth but has no cached token, Hermes would initiate an interactive browser flow during startup — blocking the main thread for up to 120 seconds and potentially deadlocking the MCP event loop via a blocking input() call if the browser callback timed out. - Replace blocking input() fallback in _wait_for_callback with TimeoutError so asyncio cancellation remains possible - Add has_valid_token() helper to check token cache before attempting any OAuth flow - In _run_http, guard OAuth setup with has_valid_token(): servers with no cached token fail fast with an actionable message instead of launching an interactive browser session at startup - Re-raise from the except block so _discover_all's asyncio.gather captures the failure and other MCP servers continue loading Fixes NousResearch#4462
684e5d9 to
961f500
Compare
teknium1
added a commit
that referenced
this pull request
Apr 3, 2026
…t loop handler, OAuth non-blocking Four fixes for MCP server stability issues reported by community member (terminal lockup, zombie processes, escape sequence pollution, startup hang): 1. MCP reload timeout guard (cli.py): _check_config_mcp_changes now runs _reload_mcp in a separate daemon thread with a 30s hard timeout. Previously, a hung MCP server could block the process_loop thread indefinitely, freezing the entire TUI (user can type but nothing happens, only Ctrl+D/Ctrl+\ work). 2. MCP stdio subprocess PID tracking (mcp_tool.py): Tracks child PIDs spawned by stdio_client via before/after snapshots of /proc children. On shutdown, _stop_mcp_loop force-kills any tracked PIDs that survived the SDK's graceful SIGTERM→SIGKILL cleanup. Prevents zombie MCP server processes from accumulating across sessions. 3. MCP event loop exception handler (mcp_tool.py): Installs _mcp_loop_exception_handler on the MCP background event loop — same pattern as the existing _suppress_closed_loop_errors on prompt_toolkit's loop. Suppresses benign 'Event loop is closed' RuntimeError from httpx transport __del__ during MCP shutdown. Salvaged from PR #2538 (acsezen). 4. MCP OAuth non-blocking (mcp_oauth.py): Replaces blocking input() call in _wait_for_callback with OAuthNonInteractiveError raise. Adds _is_interactive() TTY detection. In non-interactive environments, build_oauth_auth() still returns a provider (cached tokens + refresh work), but the callback handler raises immediately instead of blocking the MCP event loop for 120s. Re-raises OAuth setup failures in _run_http so failed servers are reported cleanly without blocking others. Salvaged from PRs #4521 (voidborne-d) and #4465 (heathley). Closes #2537, closes #4462 Related: #4128, #3436
teknium1
added a commit
that referenced
this pull request
Apr 3, 2026
…t loop handler, OAuth non-blocking (#4757) Four fixes for MCP server stability issues reported by community member (terminal lockup, zombie processes, escape sequence pollution, startup hang): 1. MCP reload timeout guard (cli.py): _check_config_mcp_changes now runs _reload_mcp in a separate daemon thread with a 30s hard timeout. Previously, a hung MCP server could block the process_loop thread indefinitely, freezing the entire TUI (user can type but nothing happens, only Ctrl+D/Ctrl+\ work). 2. MCP stdio subprocess PID tracking (mcp_tool.py): Tracks child PIDs spawned by stdio_client via before/after snapshots of /proc children. On shutdown, _stop_mcp_loop force-kills any tracked PIDs that survived the SDK's graceful SIGTERM→SIGKILL cleanup. Prevents zombie MCP server processes from accumulating across sessions. 3. MCP event loop exception handler (mcp_tool.py): Installs _mcp_loop_exception_handler on the MCP background event loop — same pattern as the existing _suppress_closed_loop_errors on prompt_toolkit's loop. Suppresses benign 'Event loop is closed' RuntimeError from httpx transport __del__ during MCP shutdown. Salvaged from PR #2538 (acsezen). 4. MCP OAuth non-blocking (mcp_oauth.py): Replaces blocking input() call in _wait_for_callback with OAuthNonInteractiveError raise. Adds _is_interactive() TTY detection. In non-interactive environments, build_oauth_auth() still returns a provider (cached tokens + refresh work), but the callback handler raises immediately instead of blocking the MCP event loop for 120s. Re-raises OAuth setup failures in _run_http so failed servers are reported cleanly without blocking others. Salvaged from PRs #4521 (voidborne-d) and #4465 (heathley). Closes #2537, closes #4462 Related: #4128, #3436
Contributor
|
Merged via PR #4757 (MCP stability fix pack). The OAuth non-blocking approach was incorporated alongside @voidborne-d's work from #4521. Thank you @heathley! |
angelburgosrosado
pushed a commit
to angelburgosrosado/hermes-agent
that referenced
this pull request
Apr 27, 2026
…t loop handler, OAuth non-blocking (NousResearch#4757) Four fixes for MCP server stability issues reported by community member (terminal lockup, zombie processes, escape sequence pollution, startup hang): 1. MCP reload timeout guard (cli.py): _check_config_mcp_changes now runs _reload_mcp in a separate daemon thread with a 30s hard timeout. Previously, a hung MCP server could block the process_loop thread indefinitely, freezing the entire TUI (user can type but nothing happens, only Ctrl+D/Ctrl+\ work). 2. MCP stdio subprocess PID tracking (mcp_tool.py): Tracks child PIDs spawned by stdio_client via before/after snapshots of /proc children. On shutdown, _stop_mcp_loop force-kills any tracked PIDs that survived the SDK's graceful SIGTERM→SIGKILL cleanup. Prevents zombie MCP server processes from accumulating across sessions. 3. MCP event loop exception handler (mcp_tool.py): Installs _mcp_loop_exception_handler on the MCP background event loop — same pattern as the existing _suppress_closed_loop_errors on prompt_toolkit's loop. Suppresses benign 'Event loop is closed' RuntimeError from httpx transport __del__ during MCP shutdown. Salvaged from PR NousResearch#2538 (acsezen). 4. MCP OAuth non-blocking (mcp_oauth.py): Replaces blocking input() call in _wait_for_callback with OAuthNonInteractiveError raise. Adds _is_interactive() TTY detection. In non-interactive environments, build_oauth_auth() still returns a provider (cached tokens + refresh work), but the callback handler raises immediately instead of blocking the MCP event loop for 120s. Re-raises OAuth setup failures in _run_http so failed servers are reported cleanly without blocking others. Salvaged from PRs NousResearch#4521 (voidborne-d) and NousResearch#4465 (heathley). Closes NousResearch#2537, closes NousResearch#4462 Related: NousResearch#4128, NousResearch#3436
02356abc
pushed a commit
to 02356abc/hermes-agent
that referenced
this pull request
May 14, 2026
…t loop handler, OAuth non-blocking (NousResearch#4757) Four fixes for MCP server stability issues reported by community member (terminal lockup, zombie processes, escape sequence pollution, startup hang): 1. MCP reload timeout guard (cli.py): _check_config_mcp_changes now runs _reload_mcp in a separate daemon thread with a 30s hard timeout. Previously, a hung MCP server could block the process_loop thread indefinitely, freezing the entire TUI (user can type but nothing happens, only Ctrl+D/Ctrl+\ work). 2. MCP stdio subprocess PID tracking (mcp_tool.py): Tracks child PIDs spawned by stdio_client via before/after snapshots of /proc children. On shutdown, _stop_mcp_loop force-kills any tracked PIDs that survived the SDK's graceful SIGTERM→SIGKILL cleanup. Prevents zombie MCP server processes from accumulating across sessions. 3. MCP event loop exception handler (mcp_tool.py): Installs _mcp_loop_exception_handler on the MCP background event loop — same pattern as the existing _suppress_closed_loop_errors on prompt_toolkit's loop. Suppresses benign 'Event loop is closed' RuntimeError from httpx transport __del__ during MCP shutdown. Salvaged from PR NousResearch#2538 (acsezen). 4. MCP OAuth non-blocking (mcp_oauth.py): Replaces blocking input() call in _wait_for_callback with OAuthNonInteractiveError raise. Adds _is_interactive() TTY detection. In non-interactive environments, build_oauth_auth() still returns a provider (cached tokens + refresh work), but the callback handler raises immediately instead of blocking the MCP event loop for 120s. Re-raises OAuth setup failures in _run_http so failed servers are reported cleanly without blocking others. Salvaged from PRs NousResearch#4521 (voidborne-d) and NousResearch#4465 (heathley). Closes NousResearch#2537, closes NousResearch#4462 Related: NousResearch#4128, NousResearch#3436
olympus-terminal
pushed a commit
to olympus-terminal/hermes-agent
that referenced
this pull request
May 16, 2026
…t loop handler, OAuth non-blocking (NousResearch#4757) Four fixes for MCP server stability issues reported by community member (terminal lockup, zombie processes, escape sequence pollution, startup hang): 1. MCP reload timeout guard (cli.py): _check_config_mcp_changes now runs _reload_mcp in a separate daemon thread with a 30s hard timeout. Previously, a hung MCP server could block the process_loop thread indefinitely, freezing the entire TUI (user can type but nothing happens, only Ctrl+D/Ctrl+\ work). 2. MCP stdio subprocess PID tracking (mcp_tool.py): Tracks child PIDs spawned by stdio_client via before/after snapshots of /proc children. On shutdown, _stop_mcp_loop force-kills any tracked PIDs that survived the SDK's graceful SIGTERM→SIGKILL cleanup. Prevents zombie MCP server processes from accumulating across sessions. 3. MCP event loop exception handler (mcp_tool.py): Installs _mcp_loop_exception_handler on the MCP background event loop — same pattern as the existing _suppress_closed_loop_errors on prompt_toolkit's loop. Suppresses benign 'Event loop is closed' RuntimeError from httpx transport __del__ during MCP shutdown. Salvaged from PR NousResearch#2538 (acsezen). 4. MCP OAuth non-blocking (mcp_oauth.py): Replaces blocking input() call in _wait_for_callback with OAuthNonInteractiveError raise. Adds _is_interactive() TTY detection. In non-interactive environments, build_oauth_auth() still returns a provider (cached tokens + refresh work), but the callback handler raises immediately instead of blocking the MCP event loop for 120s. Re-raises OAuth setup failures in _run_http so failed servers are reported cleanly without blocking others. Salvaged from PRs NousResearch#4521 (voidborne-d) and NousResearch#4465 (heathley). Closes NousResearch#2537, closes NousResearch#4462 Related: NousResearch#4128, NousResearch#3436
gweeteve
pushed a commit
to gweeteve/hermes-agent
that referenced
this pull request
Jun 2, 2026
…t loop handler, OAuth non-blocking (NousResearch#4757) Four fixes for MCP server stability issues reported by community member (terminal lockup, zombie processes, escape sequence pollution, startup hang): 1. MCP reload timeout guard (cli.py): _check_config_mcp_changes now runs _reload_mcp in a separate daemon thread with a 30s hard timeout. Previously, a hung MCP server could block the process_loop thread indefinitely, freezing the entire TUI (user can type but nothing happens, only Ctrl+D/Ctrl+\ work). 2. MCP stdio subprocess PID tracking (mcp_tool.py): Tracks child PIDs spawned by stdio_client via before/after snapshots of /proc children. On shutdown, _stop_mcp_loop force-kills any tracked PIDs that survived the SDK's graceful SIGTERM→SIGKILL cleanup. Prevents zombie MCP server processes from accumulating across sessions. 3. MCP event loop exception handler (mcp_tool.py): Installs _mcp_loop_exception_handler on the MCP background event loop — same pattern as the existing _suppress_closed_loop_errors on prompt_toolkit's loop. Suppresses benign 'Event loop is closed' RuntimeError from httpx transport __del__ during MCP shutdown. Salvaged from PR NousResearch#2538 (acsezen). 4. MCP OAuth non-blocking (mcp_oauth.py): Replaces blocking input() call in _wait_for_callback with OAuthNonInteractiveError raise. Adds _is_interactive() TTY detection. In non-interactive environments, build_oauth_auth() still returns a provider (cached tokens + refresh work), but the callback handler raises immediately instead of blocking the MCP event loop for 120s. Re-raises OAuth setup failures in _run_http so failed servers are reported cleanly without blocking others. Salvaged from PRs NousResearch#4521 (voidborne-d) and NousResearch#4465 (heathley). Closes NousResearch#2537, closes NousResearch#4462 Related: NousResearch#4128, NousResearch#3436
Egavasyug
pushed a commit
to Egavasyug/hermes-agent
that referenced
this pull request
Jun 10, 2026
…t loop handler, OAuth non-blocking (NousResearch#4757) Four fixes for MCP server stability issues reported by community member (terminal lockup, zombie processes, escape sequence pollution, startup hang): 1. MCP reload timeout guard (cli.py): _check_config_mcp_changes now runs _reload_mcp in a separate daemon thread with a 30s hard timeout. Previously, a hung MCP server could block the process_loop thread indefinitely, freezing the entire TUI (user can type but nothing happens, only Ctrl+D/Ctrl+\ work). 2. MCP stdio subprocess PID tracking (mcp_tool.py): Tracks child PIDs spawned by stdio_client via before/after snapshots of /proc children. On shutdown, _stop_mcp_loop force-kills any tracked PIDs that survived the SDK's graceful SIGTERM→SIGKILL cleanup. Prevents zombie MCP server processes from accumulating across sessions. 3. MCP event loop exception handler (mcp_tool.py): Installs _mcp_loop_exception_handler on the MCP background event loop — same pattern as the existing _suppress_closed_loop_errors on prompt_toolkit's loop. Suppresses benign 'Event loop is closed' RuntimeError from httpx transport __del__ during MCP shutdown. Salvaged from PR NousResearch#2538 (acsezen). 4. MCP OAuth non-blocking (mcp_oauth.py): Replaces blocking input() call in _wait_for_callback with OAuthNonInteractiveError raise. Adds _is_interactive() TTY detection. In non-interactive environments, build_oauth_auth() still returns a provider (cached tokens + refresh work), but the callback handler raises immediately instead of blocking the MCP event loop for 120s. Re-raises OAuth setup failures in _run_http so failed servers are reported cleanly without blocking others. Salvaged from PRs NousResearch#4521 (voidborne-d) and NousResearch#4465 (heathley). Closes NousResearch#2537, closes NousResearch#4462 Related: NousResearch#4128, NousResearch#3436
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes #4462
A failing MCP OAuth server (e.g. Atlassian) could block Hermes startup
for up to 120 seconds and permanently deadlock the MCP event loop.
Root causes:
_wait_for_callbackfell through toinput(" Code: ")on browsertimeout — a blocking syscall inside asyncio; event loop deadlocks forever
_run_httpcalledbuild_oauth_auth()unconditionally, launching abrowser flow at startup even with no cached token
Changes:
mcp_oauth.py: replaceinput()withTimeoutError; addhas_valid_token()mcp_tool.py: guard OAuth withhas_valid_token()— fail fast with"run hermes mcp auth " message; re-raise so other servers load normally
Before: Atlassian server → Hermes hangs/deadlocks
After: Atlassian server → instant warning, rest of Hermes starts normally