nemo-retriever-mcp
Skill3k starsChanged 18 days agoApache-2.0View on GitHub
Use when a task needs to search or add documents through NeMo Retriever MCP.
2 other projects use a version of this.
Trending
Quiet since May.about the whole repository
Most recently taken up by
- NVIDIA/NeMo-Retrieverword for word, first seen Oct 27, 2026
- NVIDIA/skillsword for word, first seen Oct 1, 2026
Who's talking about NVIDIA/nv-ingest
- A Hacker News story“Nvidia-Ingest: Multi-modal data extraction”Hacker News · Jan 10, 2025 · 145 points
- @github-trending.bsky.social“🚀 Skyrocketing! 🚀 (200+ new stars) 📦 NVIDIA / nv-ingest ⭐ 1,170 (+334) 🗒 Python NVIDIA Ingest is an …”Bluesky · Jan 11, 2025 · 11 likes
- @davlgd.fr“nv-ingest: an early access set of microservices for parsing hundreds of thousands of complex, messy …”Bluesky · Jan 5, 2025 · 9 likes
- NVIDIA/NeMo-Retriever“align extraction links with quickstart, NeMo-Retriever paths, and API docs”GitHub · May 15, 2026 · 1 reactions
- docling-project/docling-parse“perf: improve recursive form xobject”GitHub · Feb 11, 2026 · 1 reactions
The biggest 5 of 55 mentions linking to the repository that our crawl found. Each line is their own headline.
See all 55
Talked about on GitHub, Bluesky and Hacker News (55 times).
On GitHub · 50
- May 15, 2026align extraction links with quickstart, NeMo-Retriever paths, and API docs — NVIDIA/NeMo-Retriever1 reaction · 1 comment
- Feb 11, 2026perf: improve recursive form xobject — docling-project/docling-parse1 reaction · 2 comments
- Mar 6, 2025pdfium: use the variables rather than literal constants (signed-off variant) — NVIDIA/NeMo-Retriever1 reaction · 4 comments
- Jan 11, 2025[DOC]: Skaffold doc page non existent — NVIDIA/NeMo-Retriever1 reaction · 1 comment
- May 18, 2026docs(extraction): align extraction links for 26.03 (NVBug 6179241 batch 1) — kheiss-uwzoo/nv-ingest0 reactions
45 more on GitHub
- Mar 23, 2026Update the documents per 6000508 - point GitHub links to NVIDIA/nv-ingest — NVIDIA/NeMo-Retriever0 reactions
- Mar 11, 2026Kheiss/fix build — NVIDIA/NeMo-Retriever0 reactions
- Mar 9, 2026(retriever) change ExtractParams default to true — NVIDIA/NeMo-Retriever0 reactions
- Feb 25, 2026Add LanceDB hybrid search (dense + FTS) for retriever recall — NVIDIA/NeMo-Retriever0 reactions
- Feb 23, 2026(retriever) Add conda env + instructions to source libcudart.so.13 for OCR — NVIDIA/NeMo-Retriever0 reactions
- Feb 10, 2026Explicitly override miniforge base python version with 3.12 — NVIDIA/NeMo-Retriever0 reactions
- Feb 9, 2026Add verbose flag to HF twine upload — NVIDIA/NeMo-Retriever0 reactions
- Feb 4, 2026Check HF nightly script into default branch — NVIDIA/NeMo-Retriever0 reactions
- Feb 4, 2026Test 25.9.0 libmode CI — NVIDIA/NeMo-Retriever0 reactions · 1 comment
- Jan 29, 2026Backport Helm/Compose consistency fixes — NVIDIA/NeMo-Retriever0 reactions
- Jan 21, 2026Skip channel swapping in pdf rendering — NVIDIA/NeMo-Retriever0 reactions · 2 comments
- Dec 30, 2025fix(deps): update llama-3.2-nv-embedqa-1b-v2 to 1.10.1 — NVIDIA/NeMo-Retriever0 reactions · 2 comments
- Dec 29, 2025Ray profiling tracing — NVIDIA/NeMo-Retriever0 reactions
- Dec 22, 2025Update page elements v3 — NVIDIA/NeMo-Retriever0 reactions · 2 comments
- Dec 17, 2025Fixed summarization UDF to work with latest upstream changes — NVIDIA/NeMo-Retriever0 reactions · 1 comment
- Dec 15, 2025Bump otel-collector version in docker-compose — NVIDIA/NeMo-Retriever0 reactions
- Dec 15, 2025Add missing pdfium_hybrid option for page-elements-v3 — NVIDIA/NeMo-Retriever0 reactions
- Dec 12, 2025Fix OTEL OTLP endpoint in docker-compose — NVIDIA/NeMo-Retriever0 reactions
- Dec 11, 2025Package testing framework as Python project — NVIDIA/NeMo-Retrieversays it tested it0 reactions
- Dec 4, 2025Updating NR OCR GA container to default — NVIDIA/NeMo-Retriever0 reactions · 1 comment
- Dec 2, 2025Add documentation for pdf extraction methods — NVIDIA/NeMo-Retriever0 reactions · 1 comment
- Nov 13, 2025Update Attu service to v2.5.3 — NVIDIA/NeMo-Retriever0 reactions · 2 comments
- Nov 3, 2025[REVIEW] Increase retry count — NVIDIA/NeMo-Retriever0 reactions
- Oct 30, 2025[DOC]: Add docs for async ingestor API jobs — NVIDIA/NeMo-Retriever0 reactions · 1 comment
- Oct 24, 2025docs: Fix 2 bugs: Blackwell edition and VLM link — NVIDIA/NeMo-Retriever0 reactions
- Oct 17, 2025Fix LLM UDF Bug — NVIDIA/NeMo-Retriever0 reactions
- Oct 15, 2025remove openai requirement for embedding — NVIDIA/NeMo-Retriever0 reactions · 1 comment
- Oct 14, 2025Update UDF docs — NVIDIA/NeMo-Retriever0 reactions · 2 comments
- Oct 10, 2025Enable auth token fallback from the NGC API key to the NVIDIA API key — NVIDIA/NeMo-Retriever0 reactions
- Oct 7, 2025Add missing arg in docker publish ci workflow — NVIDIA/NeMo-Retriever0 reactions
- Oct 6, 2025[FEA]: Map Triton 503 "exceeds maximum queue size" errors to HTTP 429 — NVIDIA/NeMo-Retriever0 reactions
- Sep 29, 2025Update e2e Examples in Scripts Directory — NVIDIA/NeMo-Retriever0 reactions · 3 comments
- Sep 16, 2025[FEA]: Change Prompt for VLM in Image caption — NVIDIA/NeMo-Retriever0 reactions · 6 comments
- Aug 13, 2025add IngestService to context manager — NVIDIA/NeMo-Retriever0 reactions · 2 comments
- Jul 22, 2025Update to nvingest client version 25.6.3 — NVIDIA-AI-Blueprints/rag0 reactions
- Jul 12, 2025[BUG]: Lazy load milvus — NVIDIA/NeMo-Retriever0 reactions · 4 comments
- Jul 8, 2025Unused dependencies added to nv-ingest-client requirements.txt — NVIDIA/NeMo-Retriever0 reactions
- Jun 26, 2025Add `scipy-stubs` as dependency — NVIDIA/NeMo-Retriever0 reactions · 1 comment
- May 21, 2025Add multiple text-like filetypes — NVIDIA/NeMo-Retriever0 reactions · 2 comments
- May 4, 2025set recreate default to false — NVIDIA/NeMo-Retriever0 reactions · 2 comments
- Apr 29, 2025[BUG]: Correct name in helm chart values.yaml file - paddleocr-nim.deployed: true — NVIDIA/NeMo-Retriever0 reactions
- Apr 8, 2025[FEA]: Ingestion with custom metadata field — NVIDIA-AI-Blueprints/rag0 reactions · 2 comments
- Feb 20, 2025[BUG]: ImportError: /lib/x86_64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.32' not found — NVIDIA/NeMo-Retriever0 reactions
- Feb 11, 2025[DOC]: Fix TODO in Contributing.md — NVIDIA/NeMo-Retriever0 reactions
- Jan 30, 2025[BUG]: Failure in the langchain multimodal rag example — NVIDIA/NeMo-Retriever0 reactions
On Bluesky · 4
- Jan 11, 2025🚀 Skyrocketing! 🚀 (200+ new stars) 📦 NVIDIA / nv-ingest ⭐ 1,170 (+334) 🗒 Python NVIDIA Ingest is an early access set of microservices… — @github-trending.bsky.social11 likes and reposts
- Jan 5, 2025nv-ingest: an early access set of microservices for parsing hundreds of thousands of complex, messy unstructured PDFs and other enterprise… — @davlgd.fr9 likes and reposts
- Nov 28, 2024We are living in an era where a GPU is needed to parse a PDF. github.com/NVIDIA/nv-in... We have come a long way (slower) since John… — @docs2info.bsky.social1 likes and repost
- Jan 4, 2025More #LLM news NVIDIA released ingest, a microservice designed for extracting content and metadata efficiently from thousands of documents… — @4iamave.bsky.social0 likes and reposts · 1 reply
On Hacker News · 1
- Jan 10, 2025Nvidia-Ingest: Multi-modal data extraction145 points · 45 comments
Other versions2 projects use a version of this.
Word for word · 2
- NVIDIA/skills★ 3.9k100% same
- NVIDIA/NeMo-Retriever★ 3k100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository3 other files it gives its agents.
- AGENTS.mdAGENTS.md
- docs/AGENTS.mdAGENTS.md
- nemo-retrieverSkill
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 2 are distinctive, and 6 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 9
Use the NeMo Retriever MCP tools exposed under retriever instead of the retriever CLI.
- Very specific: line 18
Use query to search documents already available to Retriever.
- Distinctive: line 19
Never ingest documents that are already searchable, including documents in a prebuilt index.
- Very specific: line 21
Use ingest_documents only when required documents are not yet searchable and server-readable paths are explicitly available.
- Very specific: line 24
After ingestion completes, use query to retrieve evidence from the added documents.
- Very specific: line 26
Start with top_k=5 and format="hits".
- Distinctive: line 28
Read evidence hits from the tool response and ground your answer in them.
- Very specific: line 32
If query fails, inspect the error and retry with corrected parameters or a shorter equivalent query.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
NeMo Retriever MCP
Use the NeMo Retriever MCP tools exposed under retriever instead
of the retriever CLI.
The server exposes MCP tools, not MCP resources. Do not call
list_mcp_resources to discover documents or decide whether Retriever is
available; an empty MCP resource list is expected.
Workflow
- Use
queryto search documents already available to Retriever. - Never ingest documents that are already searchable, including documents in a prebuilt index.
- Use
ingest_documentsonly when required documents are not yet searchable and server-readable paths are explicitly available. Do not assume that paths visible to the agent are also readable by the Retriever server. - After ingestion completes, use
queryto retrieve evidence from the added documents. - Start with
top_k=5andformat="hits". Request more only when the first successful response does not contain enough evidence to answer the task. - Read evidence hits from the tool response and ground your answer in them.
- Preserve available document identifiers such as
document_id,source_id, orpdf_basename, along with page numbers, when the task requests citations or structured output. - If
queryfails, inspect the error and retry with corrected parameters or a shorter equivalent query. If it still fails, report the error before using another retrieval method.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

