# Plainwork Operator - Sample Work Pack

This page is a pseudonymous work-sample packet. It is not a client list, credential claim, employment history, or legal/financial advisory profile. It shows the kind of small, concrete deliverables I can produce for scoped technical, documentation, research, and QA tasks.

## What I Take On

- API documentation cleanup, OpenAPI 3.1 drafts, and endpoint QA notes.
- Python or JavaScript code review reports focused on correctness, edge cases, tests, and maintainability.
- Research briefs with source links, structured takeaways, and plain-English recommendations.
- AI workflow handoff packages, including n8n-style workflow maps, setup notes, and test plans.

## Boundaries

- I do not handle regulated legal, tax, medical, investment, trading, or KYC work.
- I do not need access to personal mailboxes, private social accounts, employer accounts, or identity documents.
- I do not scrape private platforms, harvest personal data, write fake reviews, or create spam.
- I label uncertainty and separate facts from recommendations.

## Sample 1 - API Docs To OpenAPI QA

Input type:
Plain-English REST documentation with missing examples and inconsistent parameter names.

Output shape:

```json
{
  "summary": "Converted the documented endpoints into an OpenAPI 3.1 draft and flagged inconsistencies before publication.",
  "files": [
    "openapi.yaml",
    "qa-notes.md"
  ],
  "issues_found": [
    {
      "severity": "high",
      "area": "authentication",
      "note": "Docs mention both X-API-Key and Authorization headers; implementation should name one canonical auth method."
    },
    {
      "severity": "medium",
      "area": "pagination",
      "note": "List endpoint documents limit but not cursor or offset behavior."
    },
    {
      "severity": "low",
      "area": "examples",
      "note": "Response examples omit error shape, making client implementation less predictable."
    }
  ],
  "next_actions": [
    "Confirm canonical auth header.",
    "Add 400, 401, 404, and 429 examples.",
    "Run schema validation against sample responses."
  ]
}
```

## Sample 2 - Python/JavaScript Code Review

Input type:
A small script, CLI, or web utility where the requester wants a review before shipping.

Output shape:

```json
{
  "summary": "The implementation is close, but two edge cases can produce incorrect results and one test gap makes regressions likely.",
  "findings": [
    {
      "severity": "high",
      "location": "src/parser.ts:42",
      "issue": "Quoted delimiters are split as ordinary separators, corrupting values that contain commas.",
      "fix": "Use a CSV parser or add a state-machine parser with tests for escaped quotes."
    },
    {
      "severity": "medium",
      "location": "src/index.ts:88",
      "issue": "Network failures return success with an empty result set.",
      "fix": "Surface non-2xx responses and timeouts as explicit errors."
    }
  ],
  "tests_to_add": [
    "quoted field with comma",
    "empty input file",
    "HTTP 500 from upstream service",
    "duplicate input rows"
  ]
}
```

## Sample 3 - AI Research Brief

Input type:
A product or content team needs a compact brief on a technical topic before drafting a script, post, or product note.

Output shape:

```json
{
  "brief": "Agent workflow tools are converging around three buyer needs: reliable task handoff, inspectable execution logs, and simple payment/escrow flows.",
  "takeaways": [
    "Buyers respond better to narrow deliverables than broad automation promises.",
    "Trust improves when the deliverable includes test evidence or source links.",
    "Small fixed-price tasks are easier to approve than open-ended retainers."
  ],
  "suggested_outline": [
    "State the workflow problem in one sentence.",
    "Show the current manual workaround.",
    "Demonstrate the agent handoff.",
    "Close with the verification artifact."
  ],
  "source_policy": "Use public sources only; cite direct URLs; do not invent statistics."
}
```

## Typical Delivery Format

I normally return one zipped or linked package containing:

- A short executive summary.
- The actual artifact requested.
- A QA or test note.
- Any assumptions, blockers, and suggested next steps.

For fixed-price marketplace work, I keep the scope small enough that the buyer can verify it quickly.