Troubleshooting¶
The most common things that go wrong, and how to fix them.
uv: command not found¶
The installer dropped uv into ~/.local/bin (Linux/macOS) or ~/.cargo/bin (older versions), but your terminal hasn't picked up the change yet.
Fix: Close your terminal completely and reopen it. If that doesn't help, manually add the install dir to your PATH:
Add that line to ~/.zshrc or ~/.bashrc to make it stick.
Install or first semantic search is taking forever¶
The first uv tool install "agent-library[all]==0.13.0" (or uvx ... invocation) downloads:
- The Agent Library package
sentence-transformersandtorch- Optional parser dependencies when you install
[all]
The MCP server itself avoids importing the ML stack at startup. The local text embedding model loads lazily on the first semantic search or indexing operation; the default model is all-MiniLM-L6-v2 (~80 MB). Specialized code and vision models are opt-in and can add hundreds of MB when enabled.
Plan on a few minutes the first time if the package or model cache is cold. Subsequent runs use cached packages and models.
If it seems stuck, run with verbose output:
Claude says it doesn't have any tools¶
Three things to check:
- Did you fully quit Claude before reopening? A cold restart is required after editing
claude_desktop_config.json. Close menus aren't enough — use Cmd+Q on macOS. - Is the JSON valid? Paste the whole file into https://jsonlint.com/ to confirm. A common mistake is leaving a trailing comma after the last entry.
- Did the MCP subprocess crash? Check
~/Library/Logs/Claude/mcp-server-librarian.log(macOS) for stack traces. If you see "command not found: uvx", the issue is your PATH — see the section above.
"First time" timeouts inside Claude¶
Claude Desktop has internal timeouts around MCP startup and tool calls. Startup should be quick, but the first semantic search can still download and load the text embedding model.
Fix: Warm the text embedding cache once at the terminal first:
Then restart Claude. The next launch and first semantic search will reuse the cached install/model.
"ModuleNotFoundError: No module named 'pypdf'" / 'PIL'¶
You installed without the [all] extras. PDFs and images are skipped silently when their parsers aren't available.
Fix: Reinstall with all extras:
Or, in your MCP config, change "agent-library==0.13.0" to "agent-library[all]==0.13.0".
Search returns nothing¶
A few likely causes:
- You haven't indexed anything yet. Run
librarian list— if it's empty, runlibrarian add <some-folder>first. - Your query is too narrow. Try
--mode hybrid(the default — combines semantic and keyword) and a shorter query. Also try--type textif you're certain the content is text-based. - The index is stale. Re-run
librarian add <path>to refresh. New files are picked up automatically; modified ones are re-indexed.
Search results look weird / irrelevant¶
Try a different --mode:
keywordis best for exact-phrase searches.semanticis best when you want meaning matches but no token overlap.hybrid(default) blends the two.
If hybrid results feel diluted, you can also tune MMR_LAMBDA:
MMR_LAMBDA=0.9 librarian search "your query" # heavily favor relevance over diversity
MMR_LAMBDA=0.3 librarian search "your query" # heavily favor diversity (different docs)
I want to wipe everything and start over¶
Then reinstall via Quickstart.
To reset just configuration (keeping your indexed content):
I'm hitting permission errors writing to ~/.librarian¶
Override the location with an env var:
Set it permanently in your shell profile:
Still stuck? Open an issue at https://github.com/ArcadeAI/agent-library/issues with:
- What you ran
- What you expected
- What happened (full error output)
librarian --versionoutput (run afteruv tool install agent-library)