API and agent docs

Pick a starting point

• Read selected records

  Seven read-only REST routes return a profile, work contexts, projects and capabilities, each with its sources and limits attached.

  Make a first request

• Connect an agent

  Add the MCP server to Claude, Claude Code or another client, then ask it about the published case studies and practice notes.

  Connect a client

• Fetch the plain text

  llms.txt, llms-full.txt and a Markdown version of each case study, for anything that reads text.

  See the files

The REST API and the MCP server are separate. REST serves a small selection of records, chosen and qualified one by one. MCP and the text files cover the whole published portfolio. Similar names do not mean the same fields or the same scope.

Your first request

There is no key to get. Send a GET with an Accept header and read the answer.

curl -s -H "Accept: application/json" https://jason.sonderman.info/api/v1/profile

{
  "data": {
    "capability_ids": [
      "cap_leadership",
      "cap_research",
      "cap_systems"
    ],
    "experience_ids": [
      "exp_arcos",
      "exp_cerner",
      "exp_tvh"
    ],
    "id": "pro_jason",
    "name": "Jason Sonderman",
    "qualifications": [
      {
        "kind": "scope",
        "text": "Scoped professional account; no unreviewed employment titles, tenure, credentials or contact fields."
      }
    ],
    "revision": 1,
    "sources": [
      {
        "availability": "not_publicly_available",
        "id": "src_controltoweraccount",
        "kind": "account",
        "label": "Jason Sonderman retrospective account, with the required statement-specific limitations",
        "public_url": null,
        "qualifications": [
          {
            "kind": "uncertainty",
            "text": "Retrospective account; original employer records not independently inspected. Private source material is not provided by this API."
          }
        ],
        "revision": 1,
        "type": "source_reference"
      },
      {
        "availability": "not_publicly_available",
        "id": "src_learningjourneysaccount",
        "kind": "account",
        "label": "Jason Sonderman retrospective account, with the required statement-specific limitations",
        "public_url": null,
        "qualifications": [
          {
            "kind": "uncertainty",
            "text": "Retrospective account; original employer records not independently inspected. Private source material is not provided by this API."
          }
        ],
        "revision": 1,
        "type": "source_reference"
      },
      {
        "availability": "not_publicly_available",
        "id": "src_upliftaccount",
        "kind": "account",
        "label": "Jason Sonderman retrospective account, with the required statement-specific limitations",
        "public_url": null,
        "qualifications": [
          {
            "kind": "uncertainty",
            "text": "Retrospective account; original employer records not independently inspected. Private source material is not provided by this API."
          }
        ],
        "revision": 1,
        "type": "source_reference"
      }
    ],
    "type": "profile"
  },
  "meta": {
    "api_version": "1",
    "revision": "r_5dd98d79d429a73f252b7d8b2205b657b8962e9a5ff157d87fbacfb72e2ec638",
    "coverage": "Selected qualified professional retrospective accounts: four selected studies, nineteen resources and seven assertions. Owner-approved selected content; no independent verification, media, raw sources or deployment authority.",
    "example": false
  },
  "links": {
    "self": "/api/v1/profile",
    "profile": "/api/v1/profile",
    "experience": "/api/v1/experience",
    "projects": "/api/v1/projects",
    "capabilities": "/api/v1/capabilities",
    "openapi": "/api/openapi.json"
  }
}

Next, follow a link to a project or read the full REST reference.

Limits at a glance

What you can explore

The current selection contains a professional profile, three scoped work contexts, and four case-study accounts. Capabilities connect to specific work and supporting account-based evidence.

• ARCOS Control Tower Field Research covers UX field-research leadership. Leadership does not mean sole research authorship.
• TVH Uplift Design System covers Jason’s initial setup of design tokens, shared values used by interfaces, and his scoping of where their source files would live. Conor Sixta’s later completion and governance work is credited separately. The TVH work context also credits shared governance with Engineering Director Jennifer Dennon.
• Cerner Learning Journey Portal, the Healthcare Learning Management System pilot, covers initial discovery with product lead Megan McClukin. The work context preserves the shared research origins and Jason Wander’s vision.
• Harmony design system covers Jason’s architecture leadership and hands-on contribution at Arcos, including work with a senior designer on the token architecture. It also carries selected statements about how token descriptions and chip guidance explain intended use. Other contributors owned or completed parts of the component and implementation work, so the account does not claim sole authorship, universal adoption or ownership of every component.

What you might use it for

A recruiter could find a relevant example and prepare a more specific interview question. A hiring manager could follow a capability to the work used to support it and see who contributed. A developer could reuse the selected structured information in an authorized application. An AI agent with authorized access could summarize the records while retaining their attribution and limitations.

These are intended uses. They are not claims about measured recruiting results, application adoption or AI-answer accuracy.

How to read the information

The professional history was reviewed with Jason. The source descriptions identify retrospective accounts; original employer records were not independently inspected. A supporting evidence link describes the basis for a particular statement. It does not make that statement independently verified or establish sole authorship.

Capability links are editorial associations with selected examples, not proficiency ratings. A source with no public URL means the original material is not supplied through this API. Missing dates, titles, contact details and outcomes should remain unknown rather than be guessed.

Keep the explanation and limitations with any summary you reuse. The selection is a small, read-only view of professional information; it does not provide a complete career history. Images, original working files, private records and additional studies are outside the current selection.

Choose the correct environment

No unattended login, cookie export, bearer/OAuth key or administration-token bypass is defined for the protected preview. Do not copy credentials. OpenAPI retains https://portfolio.example as a fictional design origin and Avery examples; override the base URL with the origin you have confirmed. Real TEXT19 data and fictional Avery fixtures are separate modes and must not be combined. An earlier private staging project, https://jason-ux-stage1-rest.netlify.app, is historical and not part of this release.

Shared scope and limitations

TEXT19 has nineteen distinct resources and seven qualified assertions across four selected studies. Historical confirmation is not independent inspection of employer records. Keep attribution, collaborator credit, evidence/source limits and qualifications with every explanation. Capabilities are editorial associations, not proficiency or hiring scores. Media is empty; null original-source URLs remain unknown.

Private job-search records and derivatives, raw editors, original employer evidence and private governance/approval records are excluded. CORS, no-store and hashes do not grant rights or prove global withdrawal. The developer guide explains the current revision header, conditional rate limiting and hold/redeploy behavior. Accessibility issues and unverified evidence are recorded for after Phase 2; no conformance or new audit pass is claimed.

What the project demonstrates

The API is a working example of approaching an unfamiliar technical discipline through product-design thinking. Jason began with the audiences, the questions they would ask and the information boundaries. Those needs shape the information architecture: how records and their relationships are organized. Implementation turns that structure into a working service, and documentation explains how to use it. Responses are checked against the selected records.

AI-assisted technical collaboration supported implementation. Jason’s responsibility was to define requirements, question assumptions, evaluate tradeoffs and decide what evidence would show that the approach worked. This illustrates a way of learning through delivery; it does not establish sole code authorship or a measured learning-speed or product-impact claim.