# Pinterest Search

> Route and run bounded public Pinterest image search that turns a keyword into normalized image, pin, and title records for moodboards, visual references, and creative inspiration.

- Skill: `postplusai/pinterest-search` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add postplusai/pinterest-search`
- Raw SKILL.md: https://api.skillmd.com/api/skills/postplusai/pinterest-search/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Research & Search
- Author: postplusai (https://skillmd.com/u/postplusai)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/postplusai/pinterest-search

---


# Pinterest Search

Use this skill when the user wants public Pinterest image evidence — image
addresses, moodboard source lists, or visual inspiration links — starting from a
keyword.

Apply shared rulebook and user-guidance rules from `postplus-shared`.
When a supported command completes but evidence is empty, sparse, noisy,
off-topic, or the wrong record type, apply the `postplus-shared` reference
`research-quality-recovery.md`; hard execution errors still fail fast.
Apply `references/shared-contract.md` first, then `references/search.md` for the
keyword discovery workflow.

## Job

Turn one keyword into a bounded set of Pinterest images. Return each result
normalized to `{ image_url, pin_url, title }`, deduplicated by image address.
Run the smallest first pass that can answer the request, then stop and report
scope, count, strongest results, and next action.

## Reference Index

| User asks for | Apply |
| --- | --- |
| Any Pinterest image search from a keyword | `references/shared-contract.md`, then `references/search.md` |
| Broader pass: more results, filter variation, or a second keyword | `references/search.md` |
| Board scrape, profile scrape, pin-URL scrape, shopping/product fields, engagement metrics | Not supported on the current public surface. Say so and stop |
| Non-Pinterest image sources | Hand off; run only the Pinterest lane here |

## First Question

Ask one question only when the answer changes the route, first-pass scope, or
output shape.

| Missing | Ask |
| --- | --- |
| Keyword | `What keyword should I search Pinterest for?` |
| Too broad | `Which one keyword matters most for the first pass?` |
| Filter intent | `Should I include all pins, or only video pins?` |

Do not ask the user for credentials, implementation choice, schema fields,
hidden filters, or retry strategy.

## Run Discipline

1. Apply `references/shared-contract.md`.
2. Apply `references/search.md`.
3. Run the narrowest collection that can answer the first pass; start at the
   minimum limit of 20.
4. Normalize output to `{ image_url, pin_url, title }` and deduplicate by image
   address.
5. Stop after the first pass and report scope, count, strongest results, limits,
   and next action.

The result record shape for the route is documented in the
`postplus-shared` reference `dataset-item-schemas.md`; consult it before
writing result-processing code, and probe a single record only to verify.

Do not present a bounded first pass as the full Pinterest catalog. Use only the
public filters shown by the route.

## Public Command Boundary

- Run `postplus research run pinterest-search --query <keyword> --kind
  <all|videos> --limit <n> --wait --output <result.json>`; the minimum limit is
  20.
- Readiness diagnostics: `postplus doctor --skill pinterest-search`.
- If the owned CLI command fails, report the exact error and stop. Do not bypass
  the failure with metadata-only answers, readiness probing, local payload
  rewrites, alternate services, or unpublished tools.
- Inspect flags with `postplus research run pinterest-search --help` only when
  needed.
- Keep the first pass bounded; expand only after inspecting the first result.
- If the CLI returns a quote-confirmation challenge, run
  `postplus quote confirm --json --challenge-file <challenge.json>` and retry
  with the returned token.

<!-- BEGIN GENERATED EXECUTION EXAMPLE -->
```bash
postplus research run pinterest-search \
  --query "example topic" \
  --wait \
  --output ./result.json
```
<!-- END GENERATED EXECUTION EXAMPLE -->