peepshow/ sinks/ meilisearch

Reel #79 ◆ Hybrid search ◆

peepshow sink / meilisearch

Meilisearch — Self-hosted hybrid search — POST runs at a Meilisearch index.

Index one document per peepshow run into a Meilisearch index. Hybrid full-text + auto-embedding search; self-hosted in one Docker container.

drop · process · meilisearch

What it does

[Meilisearch](https://www.meilisearch.com) is the self-hosted hybrid search engine — full-text + auto-embedding vector search in a single binary. This sink POSTs one document per peepshow run to `/indexes/<index>/documents`, letting Meilisearch auto-create the index, manage the primary key (`run_id`), and apply any server-configured embedder. Pair with the Meilisearch dashboard to search across run transcripts, titles, and container tags by either keyword or semantic similarity, with millisecond latencies on a single VPS.

When to reach for it

  • One-binary hybrid search over peepshow run history — keyword + vector with no extra services
  • Self-host search for an LLM agent on a single VPS or a developer laptop without a managed cloud bill
  • Pair with Meilisearch's auto-embedder to get semantic search over transcripts with no client-side vector pipeline

Install

npm i -g peepshow

Use it

MEILISEARCH_URL="http://localhost:7700" \
peepshow ./demo.mp4 --sink meilisearch

Make it automatic

Register the sink once — every run fires it afterward. Scope by --when so it only runs for matching videos.

peepshow sinks add meilisearch
peepshow sinks add meilisearch --when extension=mp4,mov
peepshow sinks add meilisearch --when path=/Volumes/Work/

Configuration

  • MEILISEARCH_URL Base URL of the Meilisearch instance, e.g. `http://localhost:7700`. required
  • MEILISEARCH_INDEX Index name. Default `peepshow-runs`. Auto-created on first write.
  • MEILISEARCH_API_KEY Master / admin key. Sent as `Authorization: Bearer …`.
  • PEEPSHOW_FRAME_BASE_URL When set, the first frame URL is written to the `thumbnail_url` field.

Use with an LLM agent

Every peepshow sink reads its config from env vars and receives a single JSON payload on stdin. An LLM agent (Claude Code, Cursor, Windsurf, Gemini, Codex) can drive the Meilisearch sink automatically when three things are true:

  • the env vars below are exported in the agent's shell (or a project .env it can load),
  • the peepshow CLI is on PATH — install with npm i -g peepshow,
  • a peepshow auto-sink is registered for the run (optional but recommended — makes invocation zero-argument).

1. Set the environment

# Add to ~/.zshrc, ~/.bashrc, or a project .env the agent can load
export MEILISEARCH_URL="..."

2. Register as an auto-sink

peepshow sinks add meilisearch
peepshow sinks add meilisearch --when extension=mp4,mov

3. Example LLM session

You → drop a .mov into Claude Code.

Claude → auto-invokes /peepshow:slides ./clip.mov. peepshow extracts frames + audio, the Meilisearch sink forwards the run to the configured collection. Claude replies with a summary and a link to the created record.

The transcript text is a natural embedding target — index it next to the frames.

Write your own

A sink is any executable that reads the --emit json payload on stdin. Shell, Node, Python, Go — the spec's in docs/PLUGINS.md. Register persistent ones with peepshow sinks add-cmd 'your-command'.

Related sinks

Typesense Soft-creates the collection on first write and upserts one document per peepshow Elasticsearch POST one JSON document per peepshow run to an Elasticsearch (or OpenSearch — sam Chroma Text embeddings over the metadata + caption. Frame paths as IDs. Qdrant Named vectors, payload metadata, filter-ready. Works with Qdrant Cloud or self-h

More in Vector + AI memory

Other sink categories