Troubleshooting
The reference for when something is wrong: provider connection failures, runtime stalls, agent loop budget exhaustion, file save issues, and the log paths to inspect first.
Solutions to common issues in Team-X
Table of Contents
- Installation Issues
- Login & Account Issues
- Runtime & Agent Issues
- Provider Connection Issues
- Employee & Ticket Issues
- Budget & Billing Issues
- Performance Issues
- File & Workspace Issues
- Advanced Diagnostics
Installation Issues
Team-X installer won’t open
Symptoms:
- Double-clicking installer does nothing
- Installer window opens but freezes
- “Corrupted installer” error message
Solutions:
-
Verify download:
Check file size matches download page Re-download from github.com/Git-Rocky-Stack/Team-X/releases -
Run as administrator:
- Windows: Right-click → “Run as administrator”
- macOS: Verify app from unknown developer
- System Preferences → Security & Privacy → “Open Anyway”
-
Disable antivirus temporarily:
- Antivirus may block installer
- Add Team-X to antivirus exceptions
- Retry installation
-
Check system requirements:
- Windows: Windows 10/11 (64-bit)
- macOS: macOS 11+ (Big Sur or later)
- Linux: Ubuntu 20.04+, Debian 11+
Still not working? Try the portable version (no installation required).
Installation completes but app won’t launch
Symptoms:
- Installation succeeds
- Clicking Team-X icon does nothing
- App crashes immediately on launch
Solutions:
-
Check crash logs:
- Windows:
%APPDATA%/Team-X/logs/crash.log - macOS:
~/Library/Logs/Team-X/crash.log - Linux:
~/.config/Team-X/logs/crash.log
- Windows:
-
Clear corrupted cache:
Delete cache directory: - Windows: %APPDATA%/Team-X/Cache - macOS: ~/Library/Caches/Team-X - Linux: ~/.cache/Team-X -
Update graphics drivers:
- Team-X uses GPU acceleration
- Update drivers from GPU manufacturer
- NVIDIA: nvidia.com/drivers
- AMD: amd.com/support
- Intel: intel.com/support
-
Disable GPU acceleration:
- Launch with flag:
Team-X.exe --disable-gpu - Or set environment variable:
TEAM_X_DISABLE_GPU=1
- Launch with flag:
Login & Account Issues
”Invalid email or password” error
Symptoms:
- Cannot sign in with known credentials
- Error persists after password reset
Solutions:
-
Verify email:
- Check for typos
- Ensure using correct email (not username)
-
Reset password:
- Click “Forgot Password?” on login screen
- Check email for reset link (may be in spam)
- Create new password
-
Clear stored credentials:
Delete credentials file: - Windows: %APPDATA%/Team-X/credentials.json - macOS: ~/Library/Application Support/Team-X/credentials.json - Linux: ~/.config/Team-X/credentials.json -
Check account status:
- Verify account is not suspended
- Contact support if account locked
”Account already exists” error
Symptoms:
- Trying to sign up with email returns “already exists”
- Don’t remember creating account
Solutions:
-
Reset password:
- Use “Forgot Password?” to regain access
- Email may have been used previously
-
Check for social login:
- Account may be linked to Google/GitHub
- Try “Sign in with Google” or “Sign in with GitHub”
-
Contact support:
- Verify email ownership
- Support can merge or close duplicate accounts
Two-factor authentication (2FA) issues
Symptoms:
- Cannot receive 2FA code
- Authenticator app not working
- Lost 2FA device
Solutions:
-
Check email/SMS:
- 2FA codes sent to registered email or phone
- Wait up to 5 minutes for delivery
- Check spam folder
-
Time-sync authenticator app:
- Authenticator apps require accurate time
- Enable automatic time sync on device
- Manually sync time if needed
-
Use backup codes:
- Backup codes provided during 2FA setup
- Enter backup code instead of 2FA code
- Each code can only be used once
-
Disable 2FA (if locked out):
- Contact support with identity verification
- Provide: Email, account creation date, recent activity
- Support will disable 2FA after verification
Runtime & Agent Issues
Agent stuck in “Running” state
Symptoms:
- Agent shows “Running” for > 30 minutes
- No new messages in agent stream
- Provider not responding
Solutions:
-
Check Agent Runs Panel:
- Look for error messages in stream
- Check provider status (may be down)
-
Wait longer for complex tasks:
- Large code generation: 10-30 minutes
- Data analysis: 5-20 minutes
- Architecture planning: 15-40 minutes
-
Cancel the run:
- Agent Runs Panel → [Cancel]
- Choose cancellation scope:
- “Stop after current tool” (safer)
- “Stop immediately” (may leave incomplete state)
-
Review run logs:
- Agent run log shows where agent got stuck
- Look for: timeouts, API errors, provider failures
-
Prevent future stuck runs:
- Set timeout policies: Autonomy → Runtimes → Timeout
- Enable auto-cancellation after X minutes
- Configure provider failover
Agent run failed with error
Symptoms:
- Agent stopped with error message
- Ticket shows “Failed” status
- No work completed
Common errors and solutions:
Error: “budget_exhausted”
- Cause: Per-ticket or monthly budget exceeded
- Solution:
- Approve budget override: Autonomy → Approvals
- Increase ticket budget
- Resume agent run
Error: “provider_timeout”
- Cause: Provider API not responding
- Solution:
- Check provider status page
- Retry with backup provider
- Configure failover for future
Error: “runtime_disconnected”
- Cause: External runtime (bash, node) disconnected
- Solution:
- Restart runtime: Autonomy → Runtimes → Restart
- Check runtime process (may have crashed)
- Verify runtime configuration
Error: “mcp_server_timeout”
- Cause: MCP server not responding
- Solution:
- Restart MCP server: Autonomy → MCP → Restart
- Check MCP server logs for errors
- Verify MCP server configuration
Error: “tool_execution_failed”
- Cause: Agent tool call failed (file write, bash command, etc.)
- Solution:
- Review agent run log for specific error
- Check file permissions (write access)
- Verify command syntax
- Provide feedback to agent in ticket thread
Agent produced incorrect output
Symptoms:
- Agent completed task but output is wrong
- Code has bugs
- Analysis missed requirements
Solutions:
-
Review the agent run log:
- Check if agent misunderstood requirements
- Look for missing context or unclear instructions
-
Provide corrective feedback:
In the ticket thread, write: "The component is missing the hover state mentioned in requirements. Please add hover:bg-blue-600 to the button class." -
Resume agent with feedback:
- Agent reads feedback and corrects work
- No need to create new ticket
-
Improve ticket descriptions:
- Be more specific with requirements
- Provide examples of expected output
- Include constraints and edge cases
-
Add participants for review:
- Add senior engineer as participant
- Participant reviews before agent marks done
Provider Connection Issues
”401 Unauthorized” error
Symptoms:
- Provider returns 401 error
- Agent cannot connect to Anthropic/OpenAI
Solutions:
-
Verify API key:
- Settings → Providers → [Provider] → API Key
- Regenerate key from provider dashboard
- Copy and paste key (no extra spaces)
-
Check key expiration:
- Some keys have expiration dates
- Regenerate if expired
-
Verify key permissions:
- Key must have API access
- Check provider dashboard for key permissions
”429 Rate Limited” error
Symptoms:
- Provider returns 429 error
- “Too many requests” message
- Agent unable to continue
Solutions:
-
Wait and retry:
- Rate limits reset after 1 minute (typically)
- Wait 60 seconds then retry
-
Reduce concurrency:
- Fewer simultaneous agent runs
- Stagger ticket start times
-
Upgrade provider tier:
- Higher tiers have increased rate limits
- Check provider pricing for details
-
Use multiple provider accounts:
- Add backup API keys
- Distribute load across keys
”503 Service Unavailable” error
Symptoms:
- Provider returns 503 error
- Provider status page shows outage
Solutions:
-
Check provider status:
- Anthropic: status.anthropic.com
- OpenAI: status.openai.com
-
Switch to backup provider:
- Configure failover: Settings → Providers → Failover
- Agent automatically switches to backup
-
Wait for provider recovery:
- Most outages resolved within 1 hour
- Monitor provider status page
-
Use Ollama local:
- No external dependency
- Works during provider outages
Employee & Ticket Issues
Can’t hire employee (quota exceeded)
Symptoms:
- “Employee quota exceeded” message
- Cannot create new employee
Solutions:
-
Check current employee count:
- Mission Control → Employees
- Count active employees
-
Fire unused employees:
- Remove employees not contributing
- Settings → Employees → [Employee] → Fire
-
Upgrade plan:
- Higher tiers have increased employee quotas
- Settings → Billing → Upgrade
Ticket stuck in “Blocked” status
Symptoms:
- Ticket shows “Blocked” indefinitely
- Dependencies not completing
Solutions:
-
Check dependency tickets:
- Ticket detail view → “Blocked by” section
- Verify blocking tickets are actually done
-
Resolve dependency tickets:
- Complete blocking tickets first
- Or remove invalid dependencies
-
Manually unblock:
- Ticket detail → Edit → Dependencies
- Remove blocking dependency
-
Check for circular dependencies:
- A depends on B, B depends on A (invalid)
- Break the cycle
Cannot add participant to ticket
Symptoms:
- Participant not receiving notifications
- Participant not showing in ticket
Solutions:
-
Verify participant is employee:
- Only employees can be participants
- Hire employee first if needed
-
Check employee workspace access:
- Employee must have access to workspace
- Settings → Employees → [Employee] → Workspace Access
-
Re-add participant:
- Remove and re-add participant
- Forces participant wake
Budget & Billing Issues
Budget exhausted unexpectedly
Symptoms:
- “budget_exhausted” error on active ticket
- Monthly budget exceeded early in month
Solutions:
-
Review spend breakdown:
- Autonomy → Budgets → Spend Analysis
- Identify highest-cost tickets
-
Check for runaway agents:
- Agent Runs Panel → Look for long-running agents
- Cancel if stuck
-
Review routine costs:
- Autonomy → Routines → Check routine spend
- Pause or reduce frequency of expensive routines
-
Approve budget override:
- Autonomy → Approvals → Review override request
- Approve if legitimate spend
-
Adjust budgets:
- Increase monthly budget if needed
- Set lower per-ticket limits for risky work
Charges seem too high
Symptoms:
- Spend higher than expected
- Cannot explain costs
Solutions:
-
Review Copilot cost insights:
- Copilot → Cost Insights
- Shows breakdown by employee, ticket, provider
-
Check provider costs:
- Autonomy → Budgets → Provider Costs
- Compare token usage vs. expected
-
Review agent run logs:
- Look for repeated failed attempts (retries cost money)
- Check for unnecessary tool calls
-
Optimize provider selection:
- Use cheaper models for routine tasks
- Reserve expensive models for complex work
-
Set aggressive budgets:
- Lower per-ticket limits catch overspend early
- Monthly cap prevents runaway spend
Performance Issues
Team-X running slowly
Symptoms:
- Lagging UI
- Slow panel switching
- Delayed updates
Solutions:
-
Check CPU/memory usage:
- Open Task Manager / Activity Monitor
- Team-X should use < 500MB RAM normally
- Restart if using > 2GB RAM (memory leak)
-
Close unused panels:
- Each open panel consumes resources
- Close panels not actively used
-
Clear cache:
- Settings → Advanced → Clear Application Cache
- Restart Team-X
-
Disable animations:
- Settings → Appearance → Disable Animations
- Reduces GPU usage
-
Reduce concurrent agent runs:
- Each active run consumes resources
- Limit to 3-5 concurrent runs
High memory usage
Symptoms:
- Team-X using > 2GB RAM
- System slows down when Team-X open
Solutions:
-
Restart Team-X:
- Memory leaks accumulate over time
- Daily restart recommended
-
Reduce concurrent agent runs:
- Each run holds conversation context in memory
- Limit concurrent runs
-
Clear old ticket data:
- Archive old tickets (removes from memory)
- Settings → Data → Archive Old Tickets
-
Disable unnecessary features:
- Disable Copilot if not needed (Settings → Copilot)
- Disable routine auto-execution
File & Workspace Issues
Files not saving
Symptoms:
- Agent reports file saved but file not found
- Write operations fail silently
Solutions:
-
Check approval workflow:
- Write operations may require approval
- Autonomy → Approvals → Review pending approvals
- Approve write operations
-
Check disk space:
- Verify sufficient space on target drive
- Free up space if needed
-
Check file permissions:
- Verify Team-X has write access to directory
- Check file is not read-only
- Check directory is not locked by another process
-
Review Audit Trail:
- Audit Trail → Filter by “write” operations
- Check if file operation failed
- Look for error messages
-
Manually save file:
- Copy content from agent artifact
- Manually save to target location
- Report bug if this works (approval workflow issue)
Workspace data corrupted
Symptoms:
- Crashes when opening workspace
- Missing tickets or employees
- Data inconsistencies
Solutions:
-
Restore from backup:
- Team-X creates automatic backups hourly
- Settings → Data → Restore from Backup
- Select backup point before corruption
-
Export and reimport:
- Settings → Data → Export Workspace
- Create new workspace
- Import exported data
-
Contact support:
- Provide workspace ID and corruption symptoms
- Support may be able to repair database
- Backup your data before attempting repairs
Advanced Diagnostics
Run Doctor for health check
Doctor provides comprehensive system health check:
Navigate to: Autonomy → Doctor → Run Doctor
Doctor report includes:
- Database integrity
- Recovery readiness
- Runtime posture
- Secrets status
- Provider health
- MCP health
- Budget blockers
Doctor actions:
- Run Doctor: Full health check
- Export Report: Save report for support
- Auto-Fix: Attempt to fix issues automatically
Enable debug logging
For complex issues, enable debug logging:
-
Enable debug mode:
- Settings → Advanced → Enable Debug Logging
- Restart Team-X
-
Reproduce issue:
- Perform actions that cause problem
- Debug logs capture detailed information
-
Collect logs:
- Windows:
%APPDATA%/Team-X/logs/debug.log - macOS:
~/Library/Logs/Team-X/debug.log - Linux:
~/.config/Team-X/logs/debug.log
- Windows:
-
Disable debug mode:
- Settings → Advanced → Disable Debug Logging
- Debug logs are large and slow down app
-
Share with support:
- Attach debug log to support ticket
- Include steps to reproduce issue
Reset Team-X to factory settings
Last resort: Reset Team-X (loses all local data):
Warning: This deletes all workspace data, employees, tickets, and settings. Only do this if:
- No other solutions work
- You have a backup to restore
- Starting fresh is acceptable
Steps:
-
Close Team-X completely
-
Delete application data:
- Windows: Delete
%APPDATA%/Team-X/ - macOS: Delete
~/Library/Application Support/Team-X/ - Linux: Delete
~/.config/Team-X/
- Windows: Delete
-
Reinstall Team-X:
- Download fresh installer
- Install and create new workspace
-
Restore from backup (if available):
- Settings → Data → Restore from Backup
Still Need Help?
Open a GitHub issue
Team-X is open-source and community-supported, there is no support email or hosted help desk. Bugs and reproducible issues go in the issue tracker:
Issues: github.com/Git-Rocky-Stack/Team-X/issues
Include with your report:
- Operating system and version
- Team-X version (Help → About)
- Detailed description of issue
- Steps to reproduce (if applicable)
- Relevant logs (debug.log, crash.log)
- Screenshots (if visual issue)
Community Support
- Discussions: github.com/Git-Rocky-Stack/Team-X/discussions, Q&A, ideas, show-and-tell
Documentation
- Comprehensive User Guide: Full documentation
- Quick Start Guide: Get started in 15 minutes
- FAQ: Frequently asked questions
- Scenarios: Real-world examples