We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/duquesnay/obsidian-mcp-ts'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
SAFE_USAGE.md•1.43 KiB
# Safe Usage Guidelines to Prevent Freezes
## The Problem
Claude can freeze when:
1. Processing very large responses (>10MB)
2. Self-monitoring (processes monitoring themselves)
3. Circular references in data structures
4. Team coordinator processing entire large backlogs
## Safe Patterns
### For Team Coordinator
❌ **DON'T DO THIS:**
```
"team coordinator to implement the rest of the backlog"
```
✅ **DO THIS INSTEAD:**
```
"team coordinator to implement the next 5 tasks from the backlog"
```
Or even better:
```
"What are the next 5 incomplete tasks in the backlog?"
(review the list)
"team coordinator to implement tasks T1.1 through T1.5"
```
### For Background Processes
❌ **DON'T DO THIS:**
```bash
./monitor-script.sh > output.log &
```
✅ **DO THIS INSTEAD:**
```bash
timeout 60 ./monitor-script.sh > output.log &
```
### For Long-Running Tasks
Break them into chunks:
```
"Process items 1-50"
"Process items 51-100"
"Process items 101-150"
```
## Emergency Recovery
If Claude freezes:
1. The session will be terminated after ~90 seconds
2. You'll need to restart Claude
3. Check for any background processes: `ps aux | grep claude`
## Prevention Checklist
Before running commands:
- [ ] Is the task processing a large amount of data?
- [ ] Could it create self-referential loops?
- [ ] Are you asking for "everything" or "all"?
- [ ] Is there a background process without timeout?
If yes to any, break it down or add constraints!