Use this resource with ChatGPT

Page status: reader-facing access and grounding guide. The repository's developer/operator details remain in llm/README.md.

This resource can be used through ChatGPT at three levels. The first two are convenient for readers; the third is the supported integration route for a software application. In every case, treat the book's release identity and evidence boundary as part of the answer, not as implementation detail.

Quickest route: a ChatGPT Project

Create a Project, upload the released book PDF or the chapters relevant to the question, and add the following project instructions:

Use the attached resource as the primary authority. Answer in the reader's domain language, but preserve the resource's qualifications. State the source representation, derived representation or formulation, operating state, and exactness or preservation object. Distinguish physical assets and behavioural factors from graph and matrix views. Explain the tempting short answer and why it can fail. Cite the relevant chapter, claim, or source anchor. If the resource does not establish the answer, say unsupported by this resource and identify what evidence would be needed.

Projects keep related files, chats, and instructions together. They are useful for a reader who wants to ask follow-up questions over several sessions. See Projects in ChatGPT for the current product instructions and limits.

Begin with a question that declares the audience and task, for example:

I am a power engineer. Using only this resource, explain whether a constant-admittance load belongs in the graph, in the nodal admittance matrix, or both. Separate source-model identity from formulation placement and cite the relevant sections.

Uploading a book to a conversation or Project gives ChatGPT a snapshot of the files supplied for that workspace. It is not a live connection to this repository, so check the release identity when the answer matters.

Reusable route: a custom GPT

A custom GPT can package the same approach for a class, team, or public reader experience. Configure it with:

• the curated release bundle as Knowledge;
• the answer contract above as Instructions;
• conversation starters for students, power engineers, software engineers, and mathematical modellers;
• a requirement to name the release identity and cite chapter or claim anchors; and
• an explicit unsupported response when no qualified evidence is found.

Custom GPTs use uploaded Knowledge files and configured Instructions inside ChatGPT. They are not a live synchronization mechanism: update the Knowledge files whenever a new book release changes the supported content. See GPTs in ChatGPT and Creating and editing GPTs for current configuration and sharing details.

Creating or publishing a new GPT depends on the current account and workspace permissions. If GPT creation is unavailable, use a Project with the same files and instructions; an existing shared GPT can still be used when the reader has access to it.

Do not describe a custom GPT as an independent authority. It is a reader interface over a particular resource release. Its instructions should require the same distinctions and abstention behaviour as the book.

Developer route: the grounded access service

The repository contains a model-independent corpus and a deterministic local access service. From the repository root:

python3 scripts/serve_llm_access.py --port 8787

The service provides:

• GET /healthz for availability;
• GET /v1/manifest for the corpus release identity and hashes;
• GET|POST /v1/search for ranked source records; and
• GET|POST /v1/context for an answer-oriented context packet.

The context packet is the preferred interface for an application. It carries the supported answer basis, representation and scope, qualifications, counterexamples, structured negative results, numerical pathologies, scope boundaries, and open questions, source anchors, evidence status, and an explicit unsupported or under_retrieved result when the corpus cannot support a qualified answer. The downstream model should render that packet into the user's domain language; it should not replace the packet with uncited model memory.

Compatible local clients can use the MCP adapter:

python3 scripts/mcp_llm_server.py

The local service and MCP adapter are developer interfaces. A normal ChatGPT conversation cannot call a service running on a private computer unless a separate, trusted integration makes it available. For a hosted application, load the generated corpus into a retrieval service or API vector store and retain the corpus manifest identifier and source metadata.

Recommended answer contract

Every grounded answer should contain, as applicable:

1. a direct answer supported by the resource;
2. the named representation, formulation, state, and exactness object;
3. assumptions and load-bearing qualifications;
4. the tempting shortcut and its failure consequence;
5. a translation into the reader's domain language without changing meaning;
6. chapter, claim, or source anchors and their evidence status; and
7. an unresolved boundary or an explicit unsupported status.

The answer should not call a nodal matrix “the network” without naming the source model and formulation. It should not call a graph self-loop a shunt, or assume that a load or generator has one fixed graph membership across all studies. These are precisely the shortcuts for which the retrieval layer keeps misconception contracts.

Keeping ChatGPT context synchronized

The canonical sources are the Markdown chapters, claims ledger, vocabulary registry, evidence artifacts, and release manifest. The generated corpus is a derived access artifact. Do not edit it by hand.

At each release:

1. regenerate the corpus from canonical sources;
2. run the LLM accessibility and retrieval checks;
3. publish the corpus manifest with its source commit and hashes;
4. rebuild any hosted retrieval index or custom GPT Knowledge files; and
5. test the audience prompts and high-risk misconception cases against the new release before presenting it as current.

A Project or custom GPT uploaded before step 4 is a stale snapshot. Its answers must not be described as representing the latest book release.

Choosing the route

For scientific or engineering decisions, use a grounded route and inspect the source anchors. The model's fluency is not evidence that the resource supports the conclusion.