Daily Vibe

Troubleshooting

Common issues and solutions for Daily Vibe

Troubleshooting

This guide covers common issues you might encounter with Daily Vibe and their solutions.

Installation Issues

Command Not Found

Problem: daily-vibe: command not found after installation.

Solution:

# Check if package is installed globally
npm list -g daily-vibe

# Verify npm global bin is in PATH
echo $PATH | grep -o "$(npm bin -g)"

# If not in PATH, add to your shell profile
echo 'export PATH="$(npm bin -g):$PATH"' >> ~/.bashrc
source ~/.bashrc

# Alternative: Use full path
$(npm bin -g)/daily-vibe --version

Solution:

# Check installation
npm list -g daily-vibe

# Get npm global path
npm config get prefix

# Add to PATH if needed (restart terminal after)
$env:PATH += ";$(npm config get prefix)"

# Alternative: Use full path
& "$(npm config get prefix)\daily-vibe.cmd" --version

Permission Errors During Installation

Problem: EACCES permission errors when installing globally.

Solution:

# Change npm default directory
mkdir ~/.npm-global
npm config set prefix '~/.npm-global'

# Add to PATH
echo 'export PATH=~/.npm-global/bin:$PATH' >> ~/.bashrc
source ~/.bashrc

# Install without sudo
npm install -g daily-vibe
# Install to user directory
npm install -g --prefix ~/.local daily-vibe

# Add to PATH
echo 'export PATH=~/.local/bin:$PATH' >> ~/.bashrc
source ~/.bashrc
# Use sudo (not recommended but works)
sudo npm install -g daily-vibe

Configuration Issues

API Key Not Working

Problem: Invalid API key or authentication errors.

Solutions:

  1. Verify API Key Format:

    # OpenAI keys start with sk-proj- or sk-
    # Anthropic keys start with sk-ant-
    daily-vibe config set --show
  2. Check API Key Permissions:

    • Ensure the key has appropriate permissions
    • For OpenAI: Verify billing is set up
    • For Anthropic: Check account status
  3. Test API Key Manually:

    curl https://api.openai.com/v1/models \
      -H "Authorization: Bearer $YOUR_API_KEY"
    curl https://api.anthropic.com/v1/messages \
      -H "Content-Type: application/json" \
      -H "x-api-key: $YOUR_API_KEY" \
      -d '{"model":"claude-3-sonnet-20240229","max_tokens":1,"messages":[{"role":"user","content":"test"}]}'

Configuration Not Loading

Problem: Daily Vibe not reading configuration from files.

Solution:

# Check configuration file locations
ls -la .dailyviberc.json
ls -la ~/.dailyviberc.json
ls -la .config/daily-vibe.json

# Verify JSON syntax
cat .dailyviberc.json | jq '.'

# Test with explicit config file
daily-vibe --config .dailyviberc.json config set --show

# Debug configuration loading
daily-vibe --verbose config set --show

Analysis Issues

No Data Found

Problem: "No sessions found for the specified date range"

Solution:

# Check available data sources
daily-vibe sources scan

# Verify data paths exist
ls -la ~/.claude/projects/
ls -la ~/.codex/sessions/
ls -la ~/.codex/history/

# Check file permissions
find ~/.claude -name "*.jsonl" -type f | head -5

# Try broader date range
daily-vibe analyze range --from "30 days ago" --to today

Analysis Fails with Large Datasets

Problem: Out of memory or timeout errors with large datasets.

Solutions:

  1. Reduce Date Range:

    # Analyze shorter periods
    daily-vibe analyze range --from yesterday --to today
  2. Use JSON Output:

    # JSON output is more memory efficient
    daily-vibe analyze today --json > analysis.json
  3. Adjust Configuration:

    .dailyviberc.json
    {
      "analysis": {
        "chunkSize": 5000,
        "parallelism": 2,
        "minSessionDuration": 600
      }
    }

API Rate Limiting

Problem: Rate limit exceeded errors from LLM providers.

Solutions:

  1. Reduce Parallelism:

    .dailyviberc.json
    {
      "analysis": {
        "parallelism": 1
      }
    }
  2. Use Different Model:

    # Use faster/cheaper model
    daily-vibe analyze today --model gpt-3.5-turbo
  3. Implement Delays:

    .dailyviberc.json
    {
      "llm": {
        "requestDelay": 1000
      }
    }

Data Source Issues

Claude Code Data Not Found

Problem: Claude Code sessions not being detected.

Solution:

# Check Claude Code data location
find ~ -name "*.jsonl" -path "*/.claude/*" 2>/dev/null

# Common locations
ls -la ~/.claude/projects/
ls -la ~/Library/Application\ Support/Claude/projects/

# Manually specify paths
cat > .dailyviberc.json << EOF
{
  "dataSources": {
    "claudeCode": {
      "enabled": true,
      "paths": ["~/path/to/claude/projects/**/*.jsonl"]
    }
  }
}
EOF

Codex CLI Data Not Found

Problem: Codex CLI history not being detected.

Solution:

# Check Codex CLI data locations
find ~ -name "*.jsonl" -path "*/.codex/*" 2>/dev/null

# Verify Codex CLI installation
which codex-cli || which codex

# Check configuration paths
cat ~/.codex/config.json 2>/dev/null

# Update configuration if needed
cat > .dailyviberc.json << EOF
{
  "dataSources": {
    "codexCli": {
      "enabled": true,
      "sessionPaths": ["~/.codex/sessions/**/*.jsonl"],
      "historyPaths": ["~/.codex/history/**/*.jsonl"]
    }
  }
}
EOF

VS Code Extension Data Issues

Problem: VS Code extension data not accessible.

Solution:

# Check VS Code user data
ls -la ~/Library/Application\ Support/Code/User/globalStorage/

# Look for Codex/ChatGPT extensions
find ~/Library/Application\ Support/Code -name "*codex*" -o -name "*chatgpt*"
# Check VS Code user data
ls -la ~/.config/Code/User/globalStorage/

# Look for extensions
find ~/.config/Code -name "*codex*" -o -name "*chatgpt*"
# Check VS Code user data
Get-ChildItem "$env:APPDATA\Code\User\globalStorage"

# Look for extensions
Get-ChildItem -Recurse "$env:APPDATA\Code" | Where-Object {$_.Name -like "*codex*" -or $_.Name -like "*chatgpt*"}

Output Issues

Reports Not Generated

Problem: Analysis completes but no report files are created.

Solution:

# Check output directory permissions
ls -la ./reports/
mkdir -p ./reports && chmod 755 ./reports

# Verify output path
daily-vibe analyze today --out $(pwd)/test-reports

# Check for errors
daily-vibe --verbose analyze today --out ./debug-reports

Malformed Output Files

Problem: Generated files are corrupted or incomplete.

Solution:

# Check disk space
df -h

# Verify file integrity
file ./reports/*.md ./reports/*.json

# Test with different output format
daily-vibe analyze today --json | jq '.'

# Use absolute path
daily-vibe analyze today --out $(pwd)/reports

Network Issues

Connection Timeouts

Problem: Requests to LLM APIs timing out.

Solutions:

  1. Check Network Connectivity:

    # Test OpenAI
    curl -I https://api.openai.com/v1/models
    
    # Test Anthropic
    curl -I https://api.anthropic.com/v1/messages
  2. Configure Proxy (if behind corporate firewall):

    export HTTP_PROXY=http://proxy.company.com:8080
    export HTTPS_PROXY=http://proxy.company.com:8080
  3. Increase Timeout:

    .dailyviberc.json
    {
      "llm": {
        "timeout": 60000
      }
    }

SSL/TLS Issues

Problem: SSL certificate errors.

Solution:

# Update CA certificates
# On macOS
brew install ca-certificates

# On Ubuntu/Debian
sudo apt-get update && sudo apt-get install ca-certificates

# Skip SSL verification (not recommended for production)
export NODE_TLS_REJECT_UNAUTHORIZED=0

Performance Issues

Slow Analysis

Problem: Analysis taking too long to complete.

Solutions:

  1. Optimize Configuration:

    .dailyviberc.json
    {
      "analysis": {
        "chunkSize": 8000,
        "parallelism": 4,
        "includeSystemMessages": false
      }
    }
  2. Use Faster Model:

    daily-vibe analyze today --model gpt-3.5-turbo
  3. Filter Data:

    .dailyviberc.json
    {
      "analysis": {
        "minSessionDuration": 600
      }
    }

High Memory Usage

Problem: Daily Vibe consuming too much memory.

Solutions:

  1. Reduce Chunk Size:

    .dailyviberc.json
    {
      "analysis": {
        "chunkSize": 3000,
        "parallelism": 2
      }
    }
  2. Analyze Smaller Date Ranges:

    # Instead of analyzing a week at once
    for i in {0..6}; do
      date=$(date -d "$i days ago" +%Y-%m-%d)
      daily-vibe analyze range --from $date --to $date --out "./reports/$date"
    done

Debugging Tips

Enable Verbose Logging

# Get detailed debug information
daily-vibe --verbose analyze today

# Save debug output
daily-vibe --verbose analyze today 2>&1 | tee debug.log

Test Individual Components

# Test configuration loading
daily-vibe config set --show

# Test with explicit config
daily-vibe --config .dailyviberc.json config set --show
# Test data source scanning
daily-vibe sources scan

# Check specific paths
find ~/.claude -name "*.jsonl" -mtime -7
# Test redaction rules
daily-vibe redact test "API key sk-test123 and email test@example.com"

# Test with file
echo "sk-proj-test123" > test.txt
daily-vibe redact test --file test.txt
# Quick API test with minimal data
echo '{"test": "minimal"}' | daily-vibe analyze today --json --no-redact

Getting Help

If you're still experiencing issues:

Check Version and Environment

# System info
daily-vibe --version
node --version
npm --version

# Environment info
echo "OS: $(uname -s)"
echo "Node: $(which node)"
echo "NPM: $(which npm)"

Generate Debug Report

# Create comprehensive debug report
cat > debug-report.txt << EOF
Daily Vibe Version: $(daily-vibe --version)
Node Version: $(node --version)
OS: $(uname -a)
Configuration:
$(daily-vibe config set --show)

Data Sources:
$(daily-vibe sources scan)

Error Log:
$(daily-vibe --verbose analyze today 2>&1)
EOF

Community Support

For additional help, please:

  • Check the GitHub Issues for similar problems
  • Create a new issue with your debug report
  • Join our community discussions

FAQ

Q: Can I use Daily Vibe without an internet connection?

A: Daily Vibe requires internet access to communicate with LLM APIs. However, you can:

  • Use a local LLM server with OpenAI-compatible API
  • Set up offline analysis with tools like Ollama

Q: How much does it cost to run Daily Vibe?

A: Costs depend on your LLM provider:

  • OpenAI: ~$0.01-0.05 per daily analysis (depending on data volume)
  • Anthropic: ~$0.02-0.08 per daily analysis
  • Custom APIs: Varies by provider

Q: Is my data secure?

A: Yes. Daily Vibe:

  • Only sends processed/redacted data to LLM APIs
  • Never stores your data on external servers
  • Processes everything locally before API calls
  • Includes built-in redaction for sensitive information

Q: Can I customize the analysis prompts?

A: Currently, prompts are built-in, but you can:

  • Use different models for different analysis styles
  • Modify redaction patterns
  • Post-process the JSON output with custom scripts

Next Steps

If issues persist: