This is consistently the next step after config is confirmed valid — which is exactly where BCC's diagnostics currently stop. Users have to drop to a terminal to see logs.
Proposed solution
Add a View logs action per server that opens (or tails inline in a panel) the platform-appropriate log file for that server. Pairs naturally with the spawn-test (#2).
Core: a function server_log_path(name: str) -> Path | None in bcc_core.py that returns the platform-specific path (or None if it doesn't exist yet).
GUI: View logs button per server row; opens the file in a scrollable read-only panel with auto-tail; gracefully says "no log yet" if the file doesn't exist.
Acceptance criteria
Button appears per server in the GUI.
Opens/tails the correct platform-specific log path.
Gracefully handles "no log yet."
server_log_path() has unit tests for each platform path.
Relevant code
New — pairs naturally with check_dependency/diagnostics_text in bcc_core.py
See also issue #2 (spawn-test) — the stderr from a crash is the same information as the log
## Problem
The most common "why won't my server show up" fix is manually tailing log files:
- macOS: `~/Library/Logs/Claude/mcp-server-<name>.log`
- Windows: `%APPDATA%\Claude\logs\mcp.log`
This is consistently the *next* step after config is confirmed valid — which is exactly where BCC's diagnostics currently stop. Users have to drop to a terminal to see logs.
## Proposed solution
Add a **View logs** action per server that opens (or tails inline in a panel) the platform-appropriate log file for that server. Pairs naturally with the spawn-test (#2).
- Core: a function `server_log_path(name: str) -> Path | None` in `bcc_core.py` that returns the platform-specific path (or `None` if it doesn't exist yet).
- GUI: **View logs** button per server row; opens the file in a scrollable read-only panel with auto-tail; gracefully says "no log yet" if the file doesn't exist.
## Acceptance criteria
- Button appears per server in the GUI.
- Opens/tails the correct platform-specific log path.
- Gracefully handles "no log yet."
- `server_log_path()` has unit tests for each platform path.
## Relevant code
- New — pairs naturally with `check_dependency`/`diagnostics_text` in `bcc_core.py`
- See also issue #2 (spawn-test) — the stderr from a crash is the same information as the log
the_og
added the P1 label 2026-07-02 00:50:22 -04:00
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Problem
The most common "why won't my server show up" fix is manually tailing log files:
~/Library/Logs/Claude/mcp-server-<name>.log%APPDATA%\Claude\logs\mcp.logThis is consistently the next step after config is confirmed valid — which is exactly where BCC's diagnostics currently stop. Users have to drop to a terminal to see logs.
Proposed solution
Add a View logs action per server that opens (or tails inline in a panel) the platform-appropriate log file for that server. Pairs naturally with the spawn-test (#2).
server_log_path(name: str) -> Path | Noneinbcc_core.pythat returns the platform-specific path (orNoneif it doesn't exist yet).Acceptance criteria
server_log_path()has unit tests for each platform path.Relevant code
check_dependency/diagnostics_textinbcc_core.py