Troubleshooting

Common issues and how to resolve them.

"Authentication failed"

Your stored credentials may have expired or become corrupted. Re-authenticate:

spark logout
spark login

If the issue persists, check that your system clock is accurate — token validation depends on correct timestamps. You can verify your current authentication state with:

spark whoami

"Command not found: spark"

The Spark CLI is not on your system PATH. First, verify the installation:

npm list -g @memco/spark

If the package is not listed, reinstall:

npm install -g @memco/spark

If the package is listed but the command is still not found, your npm global bin directory may not be in your PATH. Check with:

npm bin -g

Add the returned directory to your shell's PATH configuration.

"Network error"

Spark requires an HTTPS connection to the API. Check the following:

1. Internet connection — Verify you can reach other HTTPS sites.
2. Firewall or proxy — Corporate firewalls or proxy servers may block connections to api.memco.ai. Ensure HTTPS traffic to this domain is allowed.
3. VPN — Some VPN configurations interfere with TLS connections. Try disconnecting temporarily to isolate the issue.

"No recommendations found"

The knowledge network did not find solutions matching your query and tags. Try these adjustments:

• Broaden the query — Use more general terms. Instead of a very specific error message, describe the problem category.
• Remove version-specific tags — If you are using framework:next:14.2.3, try framework:next:14 or framework:next to widen the search.
• Include more error context — Copy the full error message into the query string. More context helps Spark match against known solutions.
• Check your tags — Ensure tag types and names are correct. A typo like langauge:python instead of language:python will not match.

"IDE integration not working"

If your AI coding agent is not using Spark automatically:

1. Verify configuration — Run spark status to check that your project is configured correctly.
2. Re-initialize — Run spark init to walk through the setup again. This regenerates configuration files for your detected IDE.
3. Check agent instructions — Ensure your agent's rules or system prompt includes the Spark instructions. See IDE Setup for agent-specific configuration.
4. Restart your agent — Some agents only read configuration on startup. Restart the agent session after running spark init.

"Skills installation hangs"

The interactive prompts in skills installation can cause the process to hang. Use the -y flag to skip prompts:

npx skills add memcoai/spark-cli-skills -y

Still stuck?

If none of the above resolves your issue:

• Run spark status and include the output when reporting the problem.
• Open an issue on GitHub: github.com/memcoai/spark-cli/issues (opens in a new tab)