// Receives a text question and returns its embedding vector using gemini-embedding-001 via Vertex AI. Use this when you need a semantic embedding for a question — for downstream vector search, similarity comparison, or passing to another tool.
Runs a direct SQL query against any BigQuery table and returns results as a formatted table. Use when the question requires counts, aggregations, rankings, or distributions — not when it requires reading document text. Use {project} in the SQL as a placeholder for the GCP project name.
Runs VECTOR_SEARCH against any BigQuery table that has an embedding column. Returns the top-K most semantically similar documents. Use after the embed skill has produced an embedding file. The caller is responsible for supplying the table reference, columns, and any filters.
| name | embed |
| description | Receives a text question and returns its embedding vector using gemini-embedding-001 via Vertex AI. Use this when you need a semantic embedding for a question — for downstream vector search, similarity comparison, or passing to another tool. |
Embeds a text question using gemini-embedding-001 and returns the resulting vector as a JSON array.
python ${CLAUDE_PLUGIN_ROOT}/skills/embed/scripts/embed.py --question "<your question>"
Output is a JSON array of 3072 floats printed to stdout:
[0.012, -0.034, 0.091, ...]
--question.Google Cloud SDK with active credentials:
gcloud auth application-default login
gcloud config set project <project-id>
| Symptom | Fix |
|---|---|
GCP authentication required | Run gcloud auth application-default login |
No GCP project configured | Run gcloud config set project <project-id> |
API error 401 | Re-authenticate with gcloud auth application-default login |