Human-Systems Engineering Agent

Product design, augmented

Human-Systems Engineering brings people, technology, tasks, information, and working conditions into the same product discussion. I organized HSE Agent to help teams apply that perspective across research, product strategy, design, implementation, and validation. The team supplies field inquiry, observations, interviews, subject-matter knowledge, requirements, constraints, and business context.

HSE Agent augments your ability to synthesize evidence, examine options, and carry reasoning into product decisions. It helps with analysis and artifact preparation while you direct the inquiry, judge the recommendations, and decide what happens next. Use it to connect patterns, surface gaps and risks, and consider priorities for a minimum viable product (MVP) and validation.

A skill is a reusable set of task instructions, including when it applies, the context it needs, and how to review the output. My contribution is how these practices are organized into 22 complementary skills, context profiles, and review expectations. The AI assistant uses the selected instructions and available source material; the toolkit itself does not connect to research or operational systems.

From evidence to artifacts a team can use

Outputs include research synthesis, product strategy, workflow analyses, Product Experience Blueprints, prototype direction, experiments, validation plans, requirements, backlog stories, and executive narratives. Their value depends on whether they help the team make and explain the next decision, with evidence, assumptions, recommendations, risks, and validation needs kept distinct.

Reviewed artifacts give a delivery team or AI development tool a rationale to build from. As prototype review and field validation produce new evidence, update the product direction and its linked requirements so later work retains the reasoning behind earlier decisions.

The aim is to spend less effort reconstructing context and assembling documents, leaving more attention for investigation, design, and validation. I designed the toolkit to make human-centered practices easier to apply across teams and preserve reasoning through handoffs. Assess those intended benefits in your context: examine time and review effort, source accuracy, decision usefulness, and continuity between artifacts.

How HSE Agent, PEB, and BDX fit together

Behavior-Driven Experience (BDX) provides a way to investigate behavior and identify experience improvements. The Product Experience Blueprint (PEB) connects evidence about people and work to coherent product decisions, scope, implementation, and validation. HSE Agent contains skills that help an AI assistant apply these approaches and produce related artifacts.

PEB and BDX can be used directly with a team. The toolkit supplies reusable instructions for applying them with an AI assistant; each method remains useful independently.

Replace my mining context with your own

My experiments are grounded in mining and operational technology. My workplace toolkit therefore includes a context-adaptation skill for roles, shift handoffs, field conditions, site variation, operational risk, and subject-matter review. That context helps in my work; it should not become an assumption about yours.

Replace that layer with a description of your own users, tasks, decisions, vocabulary, working conditions, data definitions, and authority. Identify the domain experts who can resolve gaps. Keep research synthesis, workflow analysis, blueprinting, backlog development, and validation skills available alongside it.

My workplace edition also includes guidance for Freeport’s Mineral Design System (MDS) 2.0 and Flutter application development. The public edition replaces that company-specific layer with a profile for your approved components, tokens, accessibility requirements, and delivery conventions. Check examples in the other skills for assumptions that may also need adaptation.

A practical workflow

1. Choose the task. State whether you need research synthesis, a workflow analysis, a blueprint, a backlog, or a validation plan, and name the decision the output should support.
2. Supply context. Provide relevant source material, users, goals, constraints, data availability, and any existing decisions. Identify which sources are authoritative and where information is incomplete.
3. Ask for traceable reasoning. Request separate observations, interpretations, assumptions, and recommendations. Require source references where available and explicit labels for unknowns.
4. Review the result. Check claims against the inputs, test whether the workflow makes sense, and resolve disagreements with the relevant practitioners or specialists.
5. Carry learning forward. Update the artifact and its linked requirements or backlog as prototype review, data investigation, and user research change the direction.

Inspect a skill in use

This example follows a volunteer handling an equipment-loan request at a library. The scenario is fictional and included in the public kit. The sample response and review show how to check an assistant’s reasoning; they do not report an actual model test.

What to inspect in an output

Check evidence: can a claim be traced to supplied material, or has an assumption been presented as a finding? Check context: does the recommendation match users, authority, environment, and actual work? Check feasibility: does it depend on unavailable data or an unconfirmed capability?

Check behavior: are failure paths, recovery, accessibility, permissions, and human review addressed where relevant? Check usefulness: does the artifact help someone make a decision, or merely restate a framework? Record unsupported recommendations and correct the instructions or inputs as appropriate.

Connecting skills across a product workflow

Use skills together when the next decision needs a different kind of analysis. Research synthesis can organize observations into needs and opportunities; workflow analysis can clarify tasks, decisions, and handoffs. A Product Experience Blueprint then connects the supported findings to capabilities, scope, dependencies, and validation.

A Product Lifecycle Progress Map summarizes current progress, key artifacts, decisions, risks and next steps for stakeholders, linking to the blueprint for deeper intent. It remains a living overview as work advances. With that blueprint reviewed, the backlog skill can help define delivery work while preserving the user need and rationale. Carry source references, open questions, and review decisions forward so an assumption does not become a confirmed requirement simply because it appears in the next artifact.

Use the smallest combination that answers the current question, and add a skill when a new analysis is needed. As implementation or research changes the evidence, update the blueprint and linked backlog together. The environment guidance below explains how skill selection works in different tools.

Human responsibility and known limits

I treat generated artifacts as drafts to inspect against the source material. AI can omit context or invent a connection even when the instructions ask for evidence. Review recommendations with practitioners and specialists who understand the work and its consequences.

The public edition received bounded behavioral smoke tests in Codex, Copilot and Claude, including lifecycle-map creation and updates. For v1.0.19, hosted Codex installation and update checks passed; the Copilot app and both tested VS Code harnesses loaded the hosted package. Public Claude v1.0.18 was behaviorally tested; the v1.0.19 branding package has not been separately uploaded and retested in Claude. These checks do not validate every skill or possible output, general reliability, or time savings. Evaluate representative tasks in your own environment, retain failures as well as successes, and record the human decisions that shaped the final artifact.

Using the guidance across tools

I developed and tested private editions of this guidance for Claude, Codex, the Snowflake-based Prospector CoWork environment used at Freeport-McMoRan, and GitHub Copilot. Public v1.0.19 provides installable Codex, GitHub Copilot and Claude editions, plus a portable Markdown library. Codex and Copilot users can add the CD-UX/hse-agent repository as a marketplace; Claude users can upload the Claude plugin ZIP. A repository-hosted marketplace is distinct from a vendor’s central directory listing.

Once installed and discovered in Codex, skills can be selected automatically when a request matches their descriptions. Use “HSE Agent,” “hse-agent,” or “hse agent” as the short name in chat: for example, “Use HSE Agent to synthesize these interview notes and identify the next research question.” These are natural-language requests, not guaranteed commands; the plugin must be installed, enabled and available to that chat. You can also name a skill when you want a particular method. Verify both approaches during setup; automatic selection depends on the request and available skills. Other environments may require choosing an agent or supplying the instructions explicitly.

Models, permissions, connected sources, and tool access differ across environments. The toolkit supplies instructions; your environment supplies those capabilities. Follow the adapter guide to check setup before using the instructions on project work.

Try a bounded task

Use examples/try-it.md in the portable library with its source notes, then compare the response with examples/review-guide.md. The library example above demonstrates the kind of review to perform. Once you can identify supported claims, proposed behavior, and unresolved questions, repeat the exercise with a small set of notes from your own work.

A useful first result is an artifact you can explain and check: what the sources support, what remains uncertain, and what decision comes next. Carry that reviewed artifact into a blueprint or backlog when the task calls for it.

What the public edition contains

The public library contains 20 general-practice skills and two replaceable context skills, along with a task router, adapter guides, a fictional exercise, and review criteria. The router helps select the smallest relevant set of skills for the task. Platform editions package that guidance for their respective clients; your research and project context remain outside the installed package.

The public edition adapts my workplace toolkit for reuse: company-specific context and design-system guidance are replaced with resources you can customize. The catalog below retains the individual skill descriptions so you can choose by the decision or task you need to support.