New: 97,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Multimodal Embedding Serving Dev

by open-edge-platform in open-edge-platform/edge-ai-libraries

Develop the Multimodal Embedding Serving microservice itself — Poetry install, run the existing tests, navigate the wrapper/registry/handler architecture, add a new model family, and build the image…

Apache-2.0Auto-check passedAI & LLM Engineering

Install Multimodal Embedding Serving Dev

skills CLI
$ npx skills add open-edge-platform/edge-ai-libraries --skill multimodal-embedding-serving-dev -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install open-edge-platform/edge-ai-libraries multimodal-embedding-serving-dev --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/open-edge-platform/edge-ai-libraries.git skills-src && mkdir -p .claude/skills && cp -r skills-src/microservices/multimodal-embedding-serving/.github/skills/multimodal-embedding-serving-dev .claude/skills/multimodal-embedding-serving-dev && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
multimodal-embedding-serving-dev
GitHub stars
168
Token cost
~1.2k tokens
SKILL.md length
388 words
Files
6 (incl. references)
Skills in repo
29
Repo updated
First seen
Licence
Apache-2.0

At a glance

Develop the Multimodal Embedding Serving microservice itself — Poetry install, run the existing tests, navigate the wrapper/registry/handler architecture, add a new model family, and build the image…

  • Works in 3 steps: docker logs -f… → curl -s localhost:9777/model/current and… → Slow first inference = lazy OpenVINO…
  • Debugging this services code
  • SKILL.md covers When to Use, Example Prompts, Reference Lookup and Environment setup, plus 5 more sections
  • Calls docker, poetry and curl

What it does

Multimodal Embedding Serving Dev is an agent skill from open-edge-platform/edge-ai-libraries. Develop the Multimodal Embedding Serving microservice itself — Poetry install, run the existing tests, navigate the wrapper/registry/handler architecture, add a new model family, and build the image from source. Use when modifying, testing, or debugging this service's code. Not for merely deploying or calling the API — that is multimodal-embedding-serving-user.

Its SKILL.md is about 1.2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 8 other files, including reference files (for example `evals/evals.json`, `example-prompts/onboard-new-model.md` and `example-prompts/update-test-cases.md`).

It sits in AI & LLM Engineering, covering Embeddings, Microservices and Creative writing and fiction. It works with Docker. The repository describes itself as: Libraries, microservices, tools, and other reference software, supporting development of performance-optimized Edge AI applications. The licence is Apache-2.0.

When your agent uses it

  • Debugging this services code
  • Tasks that involve Embeddings
  • Tasks that involve Microservices

Example prompts

  • “/multimodal-embedding-serving-dev”

Requirements

  • Python 3
  • Docker

Workflow steps

3 steps, taken from the first numbered list in SKILL.md.

  1. docker logs -f multimodal-embedding-serving — model load, conversion,
  2. curl -s localhost:9777/model/current and /model/capabilities — ground
  3. Slow first inference = lazy OpenVINO export/compile into

What it can do on your machine

Read from SKILL.md and the folder at commit 960d2e4. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • docker
    • poetry
    • curl

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use docker and curl, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Multimodal Embedding Serving Dev loads about 1.2k tokens when it runs, and up to ~2.6k if it reads all its reference files. Until then it costs about 99 tokens; SKILL.md has 388 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~99
When it runs · the whole SKILL.md, loaded when a task matches
~1.2k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~2.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from open-edge-platform/edge-ai-libraries at commit 960d2e4, republished under its Apache-2.0 licence (© open-edge-platform). 388 words, ~1,172 tokens.

Download SKILL.mdSave it as .claude/skills/multimodal-embedding-serving-dev/SKILL.md (or your agent's skills folder). This skill also uses 5 other files; get the full folder from GitHub.
name
multimodal-embedding-serving-dev
description
Develop the Multimodal Embedding Serving microservice itself — Poetry install, run the existing tests, navigate the wrapper/registry/handler architecture, add a new model family, and build the image from source. Use when modifying, testing, or debugging this service's code. Not for merely deploying or calling the API — that is multimodal-embedding-serving-user.

Multimodal Embedding Serving — Dev

Work on the service's source. This skill assumes a repo clone of edge-ai-libraries with this microservice at microservices/multimodal-embedding-serving/; if there is no clone, clone the repo first (git clone https://github.com/open-edge-platform/edge-ai-libraries.git -b main) or — if the user only wants to use the service — switch to ../multimodal-embedding-serving-user/SKILL.md. Run all commands from the microservice root.

When to Use

  • Add or modify a model-family handler (CLIP/SigLIP/MobileCLIP/CN-CLIP/Blip2/QwenText)
  • Run or extend the existing test suite
  • Navigate the wrapper → registry → handler architecture before editing
  • Build the image from source
  • Debug model load, OpenVINO conversion, or import failures

Example Prompts

Sample Problem-solving scenarios this skill handles end-to-end:

ExampleProblem it solves
onboard-new-model.mdOnboard a new embedding model family (REST + SDK)
update-test-cases.mdUpdate test cases to cover a newly onboarded model

Reference Lookup

FileLoad when…
references/source-map.mdlocating code before editing, or adding a model family (checklist inside)
references/testing.mdrunning/adding tests or smoke-testing changes

Environment setup

bash
poetry install    # Python >=3.10,<3.14

Known limitation: mobileclip and salesforce-lavis are installed only in the Docker image (see docker/Dockerfile), not via pyproject.toml. In a bare Poetry env, MobileCLIP and Blip2 handlers fail to import — develop those families against the container:

bash
export EMBEDDING_MODEL_NAME="MobileCLIP/mobileclip_s0"
source setup.sh && docker compose -f docker/compose.yaml up -d --build

Architecture in one paragraph

src/app.py (routes, input union) → src/wrapper.py (EmbeddingModel, the high-level API — also the SDK surface) → src/models/registry.py (factory) → src/models/handlers/<family>_handler.py (per-family load_model/encode_text/encode_image, PyTorch and optional OpenVINO paths). src/models/config.py holds the 19-model registry. Map + add-a-model checklist: references/source-map.md.

Show full SKILL.md (166 more words)Show less

Test / verify loop

bash
poetry run python -m unittest tests/test_path_security.py -v   # the existing suite

Coverage is thin (path security only) — for behavior changes, smoke-test via examples/server_examples.py / examples/sdk_examples.py against a small CLIP model: references/testing.md.

Build from source

bash
export EMBEDDING_MODEL_NAME="CLIP/clip-vit-b-32"
source setup.sh                                   # sourced; hard-fails without the model name
docker compose -f docker/compose.yaml build
docker compose -f docker/compose.yaml up -d

Debug a running instance

  1. docker logs -f multimodal-embedding-serving — model load, conversion, request logs (server runs --log-level debug).
  2. curl -s localhost:9777/model/current and /model/capabilities — ground truth for what's loaded.
  3. Slow first inference = lazy OpenVINO export/compile into EMBEDDING_OV_MODELS_DIR (/app/ov_models); cached afterwards in the ov-models volume.

Contribution gotchas

GotchaConsequence
src/wrapper.py is a path dependency of multimodal-dataprep (visual-data-preparation-for-retrieval/multimodal-dataprep)API changes there ripple into that service — check its usage before changing signatures
Package is built as a wheel (poetry build; packages maps src/ → multimodal_embedding_serving)keep new modules importable under the package name; SDK users import from it
INFER_BATCH_SIZE compiles OpenVINO models to a fixed batch shapehandler changes must keep the pad/split logic intact
QwenText handlers are text-only by designdon't "fix" the 400 for images; capability flags live on the handler
Every new file needs the SPDX headerCI/license scans fail otherwise

© open-edge-platform, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 5 other files (references) in microservices/multimodal-embedding-serving/.github/skills/multimodal-embedding-serving-dev of open-edge-platform/edge-ai-libraries.

  • SKILL.md
  • evals/evals.json
  • example-prompts/onboard-new-model.md
  • example-prompts/update-test-cases.md
  • references/source-map.md
  • references/testing.md

Open the folder on GitHubat commit 960d2e4

Compare with similar skills

Multimodal Embedding Serving Dev next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Multimodal Embedding Serving Dev compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Multimodal Embedding Serving Dev this skillopen-edge-platform/edge-ai-libraries168—~1.2kAutomated safety check: PassApache-2.0
Molmim NimNVIDIA/skills3.5k1 repos~1.9kAutomated safety check: NotesApache-2.0
Vss Deploy Detection Tracking 2DNVIDIA/skills3.5k1 repos~4.5kAutomated safety check: PassApache-2.0
Genmol NimNVIDIA/skills3.5k1 repos~1.4kAutomated safety check: NotesApache-2.0
Debugging Signals PipelinePostHog/posthog-foss721—~2.4kAutomated safety check: NotesMIT
Codebase Managementgiancarloerra/SocratiCode3.3k1 repos~1.8kAutomated safety check: PassAGPL-3.0

Similar skills

  • Molmim Nim

    NVIDIA/skills

    Official

    A skill your agent uses for MolMIM, NVIDIA's BioNeMo NIM microservice for small-molecule latent-space generation and optimization.

    3.5k GitHub starsUsed in 1 repo~1.9k tokens
    DevOps & CloudAuto-check: notes

  • Vss Deploy Detection Tracking 2D

    NVIDIA/skills

    Official

    A skill your agent uses when the user wants to deploy, run, debug, tear down, or call the REST API of the RTVI-CV 2D detection / tracking microservice.

    3.5k GitHub starsUsed in 1 repo~4.5k tokens
    Backend & APIsAuto-check passed

  • Genmol Nim

    NVIDIA/skills

    Official

    Generate novel drug-like molecules using the GenMol NIM microservice.

    3.5k GitHub starsUsed in 1 repo~1.4k tokens
    DevOps & CloudAuto-check: notes

  • Debugging Signals Pipeline

    PostHog/posthog-foss

    Official

    Debug the signals pipeline locally end-to-end. An agent skill from PostHog/posthog-foss.

    721 GitHub stars~2.4k tokensUpdated today
    AI & LLM EngineeringAuto-check: notes

  • Codebase Management

    giancarloerra/SocratiCode

    Set up, index, and manage SocratiCode codebase indexing. An agent skill from giancarloerra/SocratiCode.

    3.3k GitHub starsUsed in 1 repo~1.8k tokens
    AI & LLM EngineeringAuto-check passed

  • Vss Deploy Video Embedding

    NVIDIA-AI-Blueprints/video-search-and-summarization

    A skill your agent uses when deploying, operating, integrating, or customizing the VSS RT-Embed Video Embedding microservice.

    1.9k GitHub stars~2.3k tokensUpdated today
    AI & LLM EngineeringAuto-check: notes

More from open-edge-platform/edge-ai-libraries

All 29 skills in this repo

  • Time Series Analytics User

    open-edge-platform/edge-ai-libraries

    Build a new time-series analytics use case on top of the deployed Time Series Analytics microservice — bring it up with Docker Compose (from a repo clone, or by fetching the compose files from…

    168 GitHub stars~3.1k tokensUpdated today
    Auto-check passed

  • Vss Add Nest Module

    open-edge-platform/edge-ai-libraries

    Scaffolds and wires a new NestJS service/module for the Video Search & Summarization sample app's pipeline-manager using the repo's real conventions.

    168 GitHub stars~2.5k tokensUpdated today
    Auto-check passed

  • Chatqna Helm Deploy

    open-edge-platform/edge-ai-libraries

    Deploy Chat Question-and-Answer Core to Kubernetes using Helm (OpenVINO CPU, OpenVINO GPU, or Ollama), including values.yaml configuration, helm install/upgrade, deployment verification, uninstall…

    168 GitHub stars~2.3k tokensUpdated today
    Auto-check passed

  • Generate Changelog

    open-edge-platform/edge-ai-libraries

    Generates or updates CHANGELOG.md by analyzing git commit history between two branches, tags, or revisions in ANY git repository or folder.

    168 GitHub stars~3.1k tokensUpdated today
    Auto-check passed

  • Vss Deploy

    open-edge-platform/edge-ai-libraries

    Deploys and manages VSS through setup.sh and its Docker Compose overlays.

    168 GitHub stars~4.1k tokensUpdated today
    Auto-check passed

  • Vss Deploy Helm

    open-edge-platform/edge-ai-libraries

    A skill your agent uses whenever a developer needs to deploy VSS to Kubernetes, helm install VSS, configure values.yaml for VSS, or run VSS on k8s with GPU/vLLM for the…

    168 GitHub stars~3.8k tokensUpdated today
    Auto-check passed

Works with

Categories

Questions about Multimodal Embedding Serving Dev

What does Multimodal Embedding Serving Dev do?

Develop the Multimodal Embedding Serving microservice itself — Poetry install, run the existing tests, navigate the wrapper/registry/handler architecture, add a new model family, and build the image…. Multimodal Embedding Serving Dev is an agent skill from open-edge-platform/edge-ai-libraries. Develop the Multimodal Embedding Serving microservice itself — Poetry install, run the existing tests, navigate the wrapper/registry/handler architecture, add a new model family, and build the image from source.

When should I use Multimodal Embedding Serving Dev?

Multimodal Embedding Serving Dev fits situations like: debugging this services code; tasks that involve Embeddings; tasks that involve Microservices.

How do I install Multimodal Embedding Serving Dev in Claude Code?

Run `npx skills add open-edge-platform/edge-ai-libraries --skill multimodal-embedding-serving-dev -a claude-code`. Or copy the skill folder (microservices/multimodal-embedding-serving/.github/skills/multimodal-embedding-serving-dev in open-edge-platform/edge-ai-libraries) into .claude/skills/multimodal-embedding-serving-dev in your project. Claude Code loads it when a task matches its description.

How do I install Multimodal Embedding Serving Dev in Codex?

Run `npx skills add open-edge-platform/edge-ai-libraries --skill multimodal-embedding-serving-dev -a codex`. Or copy the skill folder (microservices/multimodal-embedding-serving/.github/skills/multimodal-embedding-serving-dev in open-edge-platform/edge-ai-libraries) into .agents/skills/multimodal-embedding-serving-dev in your project. Codex loads it when a task matches its description.

Can I use Multimodal Embedding Serving Dev in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add open-edge-platform/edge-ai-libraries --skill multimodal-embedding-serving-dev -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/multimodal-embedding-serving-dev, .gemini/skills/multimodal-embedding-serving-dev, .github/skills/multimodal-embedding-serving-dev and .opencode/skills/multimodal-embedding-serving-dev in your project.

What does Multimodal Embedding Serving Dev need to run?

Going by SKILL.md and its folder, Multimodal Embedding Serving Dev needs the command-line tools its instructions call (docker, poetry and curl). Our summary lists: Python 3; Docker.

Does Multimodal Embedding Serving Dev access the network?

SKILL.md contains no URLs. Its commands use docker and curl, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Multimodal Embedding Serving Dev safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Multimodal Embedding Serving Dev use?

Multimodal Embedding Serving Dev is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Multimodal Embedding Serving Dev use?

About 1.2k tokens (SKILL.md is roughly 4.7k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 1.4k tokens, read only when the agent opens those files.

What are the alternatives to Multimodal Embedding Serving Dev?

Skills that share tags, products or a category with Multimodal Embedding Serving Dev: Molmim Nim (NVIDIA/skills, 3.5k stars), Vss Deploy Detection Tracking 2D (NVIDIA/skills, 3.5k stars), Genmol Nim (NVIDIA/skills, 3.5k stars) and Debugging Signals Pipeline (PostHog/posthog-foss, 721 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Multimodal Embedding Serving Dev?

open-edge-platform (a GitHub organization) maintains it in open-edge-platform/edge-ai-libraries, which has 168 GitHub stars. The repository holds 29 skills in this directory. The repository was last updated on October 7, 2026.

Source: open-edge-platform/edge-ai-libraries on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.