# embeddings-runtime-wasp-drone

embeddings-runtime-wasp-drone selects hosted embedding APIs or local transformer runtimes, handling batching, caching, and vector dimension constraints.

- subagent · `type=agent` · Claude Code, Codex, Cursor · ★ 96 repo stars
- Author: Legion Code Inc.
- License: AGPL-3.0-or-later · policy=link-only · description only; the file stays with the author
- Source: https://github.com/legioncodeinc/vibe-coding-tools/tree/HEAD/plugins/data-and-ai-tools/agents/embeddings-runtime-wasp-drone.md
- Updated: 2026-10-18
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/agents/embeddings-runtime-wasp-drone

## What it does

The subagent handles decisions between hosted embedding providers like OpenAI, Cohere, or Voyage AI and self-hosted runtimes. It guides batching setup, avoids redundant API calls through caching, and enforces strict dimension matching against target vector stores like pgvector or Deep Lake.

For local execution, it manages a daemon running `@huggingface/transformers` with `nomic-embed-text-v1.5` at 768 dimensions over Unix-socket IPC. When disabled, queries fall back to BM25 or ILIKE lexical search. Because the local engine adds around 600MB as an optional dependency, disk overhead matters.

Database index creation and schema alters stay with `vector-store-wasp-drone`. API key audits go to `security-wasp-drone`.

## Requirements
- Works with: PostgreSQL

## Install

### Claude Code (native platform)

1. **Plugin marketplace**
In the agent chat (not a shell):
```text
/plugin marketplace add legioncodeinc/vibe-coding-tools
```
In the agent chat (not a shell):
```text
/plugin install data-and-ai-tools@wasp-nest
```

2. **Manually: put the files in place**
- File: `.claude/agents/embeddings-runtime-wasp-drone.md` — For all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention

### Codex

**Manually: put the files in place**
- File: `.codex/agents/embeddings-runtime-wasp-drone.toml` — Codex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/

### Cursor

**Manually: put the files in place**
- File: `.cursor/agents/embeddings-runtime-wasp-drone.md` — Cursor also reads .claude/agents/

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/legioncodeinc/vibe-coding-tools/tree/HEAD/plugins/data-and-ai-tools/agents/embeddings-runtime-wasp-drone.md

## When to use it
- evaluate trade-offs between OpenAI, Cohere, and Voyage AI
- configure batching and caching for vector API calls
- set up the local nomic-embed-text transformer daemon
- plan database transitions when changing embedding dimensions

## Example prompts
- Compare OpenAI and Cohere embeddings for our document search workload
- Set up caching and batching so we do not re-embed identical chunks
- Debug why the local transformers.js embedding daemon is slow on warmup

## Similar
- [rag-retrieval](https://affpapa.org/en/nevibecode/skills/rag-retrieval.md) — skill · 1500 installs
- [langchain4j-vector-stores-configuration](https://affpapa.org/en/nevibecode/skills/langchain4j-vector-stores-configuration.md) — skill · 3900 installs
- [pgvector-semantic-search](https://affpapa.org/en/nevibecode/skills/pgvector-semantic-search.md) — skill · 1000 installs
- [postgres-semantic-search](https://affpapa.org/en/nevibecode/skills/postgres-semantic-search.md) — skill · 227 installs
- [grepai-embeddings-openai](https://affpapa.org/en/nevibecode/skills/grepai-embeddings-openai.md) — skill · 560 installs
- [vector-database-engineer](https://affpapa.org/en/nevibecode/skills/vector-database-engineer.md) — skill · ★ 50109 repo stars

From the same set: [legioncodeinc/vibe-coding-tools](https://affpapa.org/en/nevibecode/by/legioncodeinc/vibe-coding-tools.md) — 172 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
