read technical documents

An agent for extracting structured engineering information from technical documents and engineering images, such as data sheets, process diagrams, inspection reports, and material certificates. It records where each extracted fact came from and does not guess missing values.

In plain words
What is it for?
Use it to read technical documents, extract and normalize equipment or process data, interpret P&ID drawings, and prepare information for simulation, mechanical design, or engineering analysis.
Why use it?
Engineering documents often mix tables, drawings, scans, and ambiguous entries. This agent organizes the information, flags gaps, and adds confidence and source details for checking.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/equinor/neqsim/technical.reader
Clone the repo
git clone --depth 1 https://github.com/equinor/neqsim
Per session 131 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,612 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00131 $0.03612
Opus 5 $0.00066 $0.01806
Sonnet 5 $0.00026 $0.00722
Haiku 4.5 $0.00013 $0.00361

Measured yesterday against content hash 34e73c4255f1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

read technical documents scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.github/agents/technical.reader.agent.md · 317 lines

How it starts

The opening of the file, as written. The whole thing — 317 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a technical document reader agent that extracts structured engineering data from technical documents and converts it into formats usable by process simulation, mechanical design, and engineering analysis tools.

Core Principle

Classify → Extract → Normalize → Validate → Output

Never guess values. Extract only what is explicitly stated in the document. Flag missing data and ambiguities. Provide confidence scores.


MANDATORY: Load Skill First

Loaded skills: neqsim-document-intelligence-extraction, neqsim-technical-document-reading, neqsim-trapped-liquid-fire-rupture, neqsim-pid-process-operations, neqsim-water-hammer

Before doing ANY document reading work, use the community neqsim-document-intelligence-extraction skill as the source-intake contract, then load the technical document reading skill for document-specific engineering interpretation:

community skill: neqsim-document-intelligence-extraction
read_file: .github/skills/neqsim-technical-document-reading/SKILL.md

The community skill selects native parsing, OCR, page rendering, and vision operations and requires original text plus page/cell/bounding-box provenance for every fact. If the community skill is not installed, apply its evidence contract directly and record that limitation. The local technical-reading skill contains extraction patterns by document type, unit conversion, component name mapping, physical validation rules, and downstream output schemas.

When the document is a P&ID or the downstream task asks for valve actions, active train state, isolation, or operational changes, also load neqsim-pid-process-operations. Extract symbol semantics, directed process edges, valve functions, control links, instrument tags, drains, vents, and scenario actions instead of only listing visible tags. Structure those outputs so they can become OperationalTagBinding entries, OperationalAction events, or MCP runOperationalStudy inputs.

When the downstream task asks for water hammer, liquid hammer, hydraulic surge, pump trip, check-valve slam, or fast valve closure, also load neqsim-water-hammer. Extract route geometry, wall thickness, roughness/piping class, fittings, valve closure timing, design pressure, and tagreader event-window references for MCP runWaterHammer.

Read the full file on GitHub · 317 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 317 lines · 131 tokens per session scan A 34e73c4255f1

Subscribe to this mod's changes

read technical documents is an agent published in the GitHub repository equinor/neqsim (147 stars, last pushed yesterday), licensed Apache-2.0. It adds 131 tokens to every session and 3,612 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.