nemo-retriever
Uses NeMo Retriever 26.8.1 with local LanceDB indexes or deployed Retriever services. Supports PDFs, images, Office files, HTML, text, audio, and video; excludes document editing and web search.
Your agent can do more. Find the skill that makes it happen.
Find your next superpower If you are an agent, refer to our llms.txt for full access.Uses NeMo Retriever 26.8.1 with local LanceDB indexes or deployed Retriever services. Supports PDFs, images, Office files, HTML, text, audio, and video; excludes document editing and web search.
Guides agents through hypothesis testing and open-ended discovery, from understanding recipes and environments to launching baselines and analyzing iterations. Uses git and TSV logs as a research ledger while preserving human oversight.
Provides operating guidance for NeMo-RL agents using a limited workspace disk and a larger /ephemeral volume. Covers large downloads, training outputs, checkpoints, shared caches, and optional secrets for W&B or Hugging Face workflows.
Covers updates to docs/index.md and docstring formatting in NeMo-RL. Applies to writing or updating documentation.
Manages durable working-session memory under the repository's session directory. Supports periodic state preservation during long-running work and recovery after disconnects, VS Code restarts, or handoffs.
Uses bump-version.ts to update version strings and opens a release pull request. After merge, tags main and pushes the tag to complete the release workflow.
Guides assistants to the NemoClaw docs MCP server and Fern documentation in Markdown. Supports questions about installation, configuration, operation, troubleshooting, security, and learning NemoClaw.
Builds single-step or multi-step pipelines for curation, translation, SFT/PEFT, pretraining, and RL alignment. Includes benchmarks, checkpoint conversion, ModelOpt optimization, environment profiles, and evaluation of checkpoints or hosted endpoints.
Creates a Markdown policy, JSON taxonomy, and drop-in inference prompts for text and multimodal Nemotron content-safety models. Maps rough requirements or existing policies to V2 categories, with custom categories and topic-following rules.
Supports public Nemotron embed and rerank retrieval recipes. Covers planning, debugging, tuning, evaluation, export, and deployment.
Routes NVIDIA Nemotron Speech, formerly Riva, tasks for ASR, TTS, and NMT NIMs. Supports build.nvidia.com and self-hosted execution.
Supports Cypher query writing, performance optimization, graph pattern matching, vector and full-text search, subqueries, and batch writes. Covers clauses, expressions, functions, and indexes, while excluding driver migrations, API changes, and database administration.
Skills give your agent reusable instructions for a specific job. Pick one, read what it does, and bring it into your workflow.
A name only tells half the story. Search the full description and instructions to find the right fit.
Read the skill, visit its source, and see exactly what you’re adding to your agent.
Copy the install command from a skill page and run it in your project.
npx skillycli add owner/repo --skill name