iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Nebula Agent is a Node.js command-line demo that answers questions from Sanity-hosted article content and flags where two sources disagree. When it finds a conflict, it shows both claims side by side with the URL of each source. The example is built around a fictional board game called Nebula, so its numbers are demonstration values, not real rules. Its main lesson is that an agent can find and display a contradiction easily, but deciding which source wins requires rules that the developer has to write.
How the agent answers a question
The project was submitted to a Sanity challenge by its author, Victor, and is described in a DEV Community post published on September 27, 2026. The chain from question to answer works like this:
- The user asks a question in plain language, for example, “According to the Nebula rulebook and errata, how many energy tokens do players start with? Are there any contradictions?”
- The language model, Gemini, writes a GROQ query against the Sanity dataset. The author states that the model is given a
query_documentsfunction for this purpose. - The function sends that GROQ query through
@sanity/clientto the Sanity Content Lake and returns the matching records as JSON. - The JSON is passed back to Gemini, which writes the final answer using
@google/genaifunction calling. - The answer is governed by a prompt that tells the model how to handle disagreement between sources (covered in the next section).
The author describes the agent as a loop that retrieves records, compares them, and returns an answer that keeps each claim tied to its source. Nothing in the write-up suggests a separate ranking service or a verification step outside the model’s own reasoning.
Why the schema has to be spelled out
The most useful practical detail in the write-up is about schema. The article records a title, slug, body, and source field, and the author reports that the model initially guessed document types and returned empty results. Naming the actual article type in the prompt fixed the problem, because the model could then build queries against real records. The author’s own summary: “The hardest part wasn’t the LLM logic, it was teaching the agent the schema.”
#1 Best Overall
- 【Leading AI Mini Workstation】MINISFORUM AI MS-S1 Max Workstation comes with AMD Ryzen AI Max+ 395 processor, which uses AMD's latest generation Zen 5 architecture. It has 16 Cores and 32 Threads, the boost clock is up to 5.1GHz. The overall processor performance is up to 126 TOPS, and the NPU performance reaches up to 50 TOPS. AMD Ryzen AI enables improved productivity, advanced collaboration, and improved efficiency.
- 【AMD Radeon 8060S Graphics 】The MS-S1 Max Mini PC equipped with AMD Radeon 8060S Graphics which built on the new generation of RDNA 3.5 architecture AMD graphics, it brings ultra-high frame rate experiences and advanced content creation features anywhere and delivers staggering performance. It can handle all your computing and multimedia tasks efficiently.
- 【Five 8K Video Output】This MS-S1 Max Workstation comes with five video outputs, 1x HDMI (8K@60Hz), 2x USB4(40Gbps,Alt DP2.0,PD out 15W) and 2x USB4 V2(80Gbps,Alt DP2.0,PD out 15W) Outputs, which support multiple monitors display at the same time and provide a larger and wider filed of view and improve your work efficiency. It is used in fields that require high-performance computing and graphics processing, including digital signage and securities trading, as well as work that uses CAD, such as engineering design, scientific calculations, animation production, and post-production for movies and television.
- 【 Fast and Stable Wire & Wireless Speed】It comes with Two 10G Lan Ports for wired connection and and Wi-Fi 7 / BT5.4 for wireless connection, which increased the network speed greatly and expand its functions and improved performance of computer to a large extent and allows you to use more networks such as software routers (OpenWRT / DD-WRT / Tomato etc.), firewalls, NAT, network isolation etc.
- 【Large Storage & Flexible Expandability】This Workstation equipped with 128GB LPDDR5-8000MHz + 2TB M.2 2280 PCIe4.0 SSD. There is another PCIe4.0 SSD slot available for up to 8TB, these SSD slots are compatible with RAID0 and RAID1, you can store movies, videos, photos, important files easily. What’s more, it also comes with 1x standard PCIex16 slot(PCIe4.0x4) inside.
An illustrative query of the kind the agent needs to write looks like this. It is a generic GROQ sketch for this schema, not a line copied from the repository:
*[_type == "article"] {
title,
"slug": slug.current,
body,
source
}The author also says the client was configured with useCdn(false) so that queries hit the Content Lake directly. That is the author’s implementation choice; the write-up does not measure latency or freshness, so treat it as a configuration detail rather than a performance claim.
The contradiction rule in the prompt
The behaviour that makes the project interesting is a short instruction in the prompt:
Rank #2
- 【Leading AI Mini Workstation】MINISFORUM AI MS-S1 Max Workstation comes with AMD Ryzen AI Max+ 395 processor, which uses AMD's latest generation Zen 5 architecture. It has 16 Cores and 32 Threads, the boost clock is up to 5.1GHz. The overall processor performance is up to 126 TOPS, and the NPU performance reaches up to 50 TOPS. AMD Ryzen AI enables improved productivity, advanced collaboration, and improved efficiency.
- 【AMD Radeon 8060S Graphics 】The MS-S1 Max Mini PC equipped with AMD Radeon 8060S Graphics which built on the new generation of RDNA 3.5 architecture AMD graphics, it brings ultra-high frame rate experiences and advanced content creation features anywhere and delivers staggering performance. It can handle all your computing and multimedia tasks efficiently.
- 【Five 8K Video Output】This MS-S1 Max Workstation comes with five video outputs, 1x HDMI (8K@60Hz), 2x USB4(40Gbps,Alt DP2.0,PD out 15W) and 2x USB4 V2(80Gbps,Alt DP2.0,PD out 15W) Outputs, which support multiple monitors display at the same time and provide a larger and wider filed of view and improve your work efficiency. It is used in fields that require high-performance computing and graphics processing, including digital signage and securities trading, as well as work that uses CAD, such as engineering design, scientific calculations, animation production, and post-production for movies and television
- 【 Fast and Stable Wire & Wireless Speed】It comes with Two 10G Lan Ports for wired connection and and Wi-Fi 7 / BT5.4 for wireless connection, which increased the network speed greatly and expand its functions and improved performance of computer to a large extent and allows you to use more networks such as software routers (OpenWRT / DD-WRT / Tomato etc.), firewalls, NAT, network isolation etc.
- 【Large Storage & Flexible Expandability】This Workstation equipped with 64GB LPDDR5-8000MHz + 2TB M.2 2280 PCIe4.0 SSD. There is another PCIe4.0 SSD slot available for up to 8TB, these SSD slots are compatible with RAID0 and RAID1, you can store movies, videos, photos, important files easily. What’s more, it also comes with 1x standard PCIex16 slot(PCIe4.0x4) inside.
- Show conflicts openly. “When two sources contradict each other, show both claims side by side with their sources.”
- Attribute every claim. “Cite the source URL for every claim.”
- Do not fill gaps. “Never invent information.”
These rules make the output auditable: a reader can see each claim, where it came from, and that the model was told not to add anything. They do not, by themselves, tell the model which claim is correct.
What the two sample answers show
The demo includes two built-in examples. Both use fictional Nebula figures, and the cited example.com URLs are placeholders the author says should be replaced in a real deployment.
| Question | Claim in one source | Conflicting claim in another source | What the reader learns |
|---|---|---|---|
| Energy tokens at the start of a game | 5 energy tokens (rulebook) | 8 energy tokens (errata) | The answer must show both values and their URLs, not silently pick one. |
| Win condition | 10 stars | 12 stars | The disagreement is surfaced, but the sample does not establish which value is authoritative beyond the labels in the sources. |
In the energy-token example, the errata is treated as authoritative because its title identifies it as errata. That is a naming convention, not a provenance check. A real system would need a rule that the model can apply consistently, which is the subject of the next section.
Rank #3
- Built for Local AI and Advanced Workflows – The BOSGAME M5 AI Mini PC is powered by AMD Ryzen AI Max+ 395 with 16 cores, 32 threads, up to 5.1GHz, 50 TOPS NPU performance and up to 126 TOPS total AI performance. It is designed for local AI inference, private AI assistants, coding, data analysis, virtualization, content creation and demanding multitasking while keeping sensitive data on the device.
- 128GB Unified Memory for Large Models and Creative Projects – M5 includes 128GB LPDDR5X-8000 unified memory, giving the CPU and Radeon 8060S graphics access to a large shared memory pool. This helps support memory-intensive AI workloads, large project files, multiple virtual machines, 3D work, video editing and complex professional applications without the capacity limits of typical 32GB or 64GB mini computers.
- Radeon 8060S Graphics for Creation, Rendering and Gaming – Integrated Radeon 8060S graphics with 40 RDNA 3.5 compute units delivers high-end visual performance without a separate graphics card. Use the M5 creator workstation for 4K video editing, 3D rendering, CAD, AI image workflows, high-resolution media and modern gaming, while maintaining a compact desktop footprint.
- 2TB PCIe 4.0 SSD and Flexible Expansion – A pre-installed 2TB NVMe PCIe 4.0 SSD provides fast access to models, datasets, media libraries and project files. A second M.2 2280 PCIe 4.0 slot allows additional storage expansion, while the SD 4.0 card reader supports efficient photo and video workflows for creators and production teams.
- Professional Connectivity and Four-Display Support – Dual USB4 ports, HDMI 2.1 and DisplayPort 1.4 support up to four displays and resolutions up to 8K@60Hz. WiFi 7, Bluetooth 5.4 and 2.5GbE deliver fast networking for cloud collaboration, NAS access and business deployment. Windows 11 Pro, performance-mode switching, Wake-on-LAN and auto power-on support flexible workstation use.
Source authority: the part the demo does not solve
Detecting that two sources disagree is the easy half. The author’s own framing is that a contradiction detector needs explicit rules for authority, because a disagreement alone does not reveal which source is right. The demo does not report a measured accuracy rate, a benchmark, a controlled evaluation, or proof that the model resolves authority reliably. If you adapt the approach, build authority into the content itself rather than relying on the model to infer it. Useful signals include:
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minute- Document type. A rulebook, an errata, and a strategy guide carry different weight. Store the type in a field such as
sourceor a dedicateddocTypevalue. - Publication date. A dated errata should be able to supersede an earlier rulebook. Record the date as a field the query can sort or filter on.
- Explicit supersession. A field that names the document a newer record replaces removes guesswork. Without it, the model is left to interpret titles.
- An owner or editorial rule. State who maintains each source and what the precedence order is, and put that order in the prompt.
Where these fields are missing, the honest behaviour is what the prompt already asks for: show both claims and say that the precedence is unclear.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Keyword search versus a schema-aware agent
The author’s argument is that a plain keyword search over the same articles would find both conflicting documents, but would not know which one is authoritative. The table below compares the approaches on the axes the write-up uses. It reflects the author’s explanation of the demo, not a measured comparison.
Rank #4
- Speed up your tasks with AI: Unlock new levels of productivity and creativity by upgrading to Intel Core Ultra processors with built-in AI.
- Supports multiple monitors: Connect up to four FHD monitors using DisplayPort and Daisy Chaining*. Or connect two 4K displays using HDMI 2.1 port and DisplayPort.
- Effortless upgrades: The tool-less entry and removable side panel let you quickly access the internal components, making upgrades convenient and stress-free.
- Ready for business: Keep your data secure with a hardware TPM security chip. And when you need to step away from your desk, simply secure your desktop using the built-in lock slot or padlock loop.
- Style meets sustainability: Dell Tower Desktop seamlessly combines elegance with sustainability. Its sleek, modern design, crafted from recycled materials and featuring refined corners, makes it a stylish addition to any home or office.
| Approach | Finds both conflicting documents | Knows the record structure | Keeps source URLs in the answer | Chooses an authoritative source |
|---|---|---|---|---|
| Keyword search over raw text | Yes, per the author’s explanation | No | Only if the search tool returns it | No |
| Schema-aware agent (Nebula Agent) | Yes, through GROQ queries against typed records | Yes, when the schema is named in the prompt | Yes, per the prompt rule | Only through rules and fields the developer supplies |
Stack and where to find the code
The author reports the following components. The versions are the names given in the write-up; they were not independently checked against current releases.
- Node.js command-line application
- Gemini 3.5 Flash Lite as the language model
@sanity/clientfor GROQ queries@google/genaifor function calling- Sanity Content Lake as the data store
The write-up links a GitHub repository. Its code snippets are the basis for the description above; a line-by-line review of the repository was not part of this article. For the author’s full account, read the original DEV Community post.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Adapting the pattern to your own content
- Define a document type with at least
title,slug,body, andsource, then add a date and a document-type field. - Name that exact type in the system prompt so the model writes queries against it rather than guessing.
- Expose a single query function that returns JSON and passes results back to the model without modification.
- Keep the “show both claims, cite every URL, invent nothing” rules, and add your precedence order directly after them.
- Test with pairs of records you already know conflict. Check that both claims and both URLs appear, and that the precedence wording matches the fields you stored.
What to take from it
Nebula Agent is a clear example of how structured content makes an AI answer traceable: queryable records, a named schema, and source URLs carried through to the output. It is not evidence that a model can judge source authority on its own. If your content has conflicting sources, the work that matters is in the fields, dates, and precedence rules you add around the model.
Several points remain unverified in the public material: the repository’s current state, the package and model versions at the time of your own build, and any accuracy figures. Check those directly before relying on the approach in production.
Source: Nebula Agent: An AI That Surfaces Contradictions in Structured Content, DEV Community, September 27, 2026.
Quick Recap
“
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Free tools Windows power users keep installed
One-click scans. No signup required.
