Spyglass

View saved version →

---
name: spyglass-accounts
description: Resolve brands and creators by name, handle, ID, or category brief, and index missing accounts into Spyglass. Use whenever a request names specific brands/creators, asks to discover competitors or influencers in a niche, or needs an account added before pulling its media, insights, campaigns, or landing pages.
---

# spyglass accounts

## Tool availability in ChatGPT

Use the tools exposed by the connected Spyglass MCP server, matching the action names used below. Tool availability and the live input schema take precedence over examples in this skill. The indexing instructions below apply only when both `addAccounts` and `getAccountsIndexStatus` are available and the user has authorized indexing. If either tool is unavailable, do not attempt indexing or claim an account was added; explain that the account is not currently indexed and continue with available accounts. Use only demographic enum values accepted by the live schema.


Find brands and creators in the Spyglass corpus with `getAccounts`, or index missing accounts on demand with `addAccounts` and `getAccountsIndexStatus` so their ads, organic posts, insights, campaigns, and landing pages become searchable.

For specific named brands or creators, always start with `getAccounts` in `brand_name` mode, verify that the returned row's `summary` matches the intended company, and call `addAccounts` if the account is missing. For discovering competitors or creators in a category, use `getAccounts` in `smart` mode with a concrete market description and demographic/follower filters.

## 1. Pick the right `getAccounts` mode

`getAccounts` has three modes (`brand_name`, `smart`, `ids`):

- **`brand_name` (exact / substring lookup on names, handles, and domains)**:
  Use whenever the user names one or more specific brands, creators, `@handles`, or domains.
  - Batch **all** named accounts into a single `brand_name` call (`brandNames` accepts up to 100 strings).
  - For each brand, include the exact name plus a spaced/unspaced variant and bare domain or handle if known (e.g., `["SotoTennis", "Soto Tennis", "sototennis.com"]`).
  - Omit `type` if you are unsure whether an account is indexed as a `BRAND` or `CREATOR`.

- **`smart` (vector discovery by market, niche, or audience)**:
  Use when discovering competitors, brands in a vertical, or creators in a niche (`"DTC electrolyte hydration powder brands"`, `"K-beauty skincare creators"`).
  - Write `query` as a self-contained market brief with concrete nouns (product type, category, audience, price tier). Never pass a vague word like `"competitors"` or `"similar brands"`.
  - When finding competitors for a known brand, resolve the reference brand first, read its `summary`, and build the `smart` query from its actual product category—then exclude the reference brand from your answer.

- **`ids` (re-hydration by ID)**:
  Use when you already have Spyglass UUIDs or numeric Facebook Ad Library IDs from prior tool calls or citations and need their profile summaries or follower counts.

## 2. Verify matches before pulling media

`brand_name` matches substrings (`ILIKE %name%`), and `smart` always returns the closest semantic neighbors even when a specific brand is not in the database.

- **Check `summary` for substring and homonym collisions:**
  - Short names frequently match unrelated accounts: `"Anessa"` matches `"Danessa Myricks Beauty"` and `"Joicy Vanessa"`; `"Chery"` matches `"Cheryl Porter"`; `"Tesa"` matches `"tastesalud"`; `"Collider"` (the non-alcoholic beer) matches `"Collider"` (`collider.com`, the entertainment news site).
  - Always read the `name`, `summary`, and `location` fields before passing an `id` to `getMedia`. Reject rows whose industry or product does not match the user's request.
- **Do not mistake `smart` fallback results for a named account:**
  - If `brand_name` returns `0` matches and you retry with `smart`, `smart` will return 20 accounts in the same category even if the

Preview of the first 4,000 characters. Read full instructions →

View saved version →

---
name: spyglass-media
description: Find paid Meta ads, organic Instagram posts, and TikTok videos with creative metadata, hooks, angles, and performance signals. Use for competitive ad research, discovering winning creatives on a topic, or pulling a brand's ad library. Supports corpus semantic search, taxonomy filtering, and brand-level retrieval.
---

# spyglass media

## Tool availability in ChatGPT

Use the tools exposed by the connected Spyglass MCP server, matching the action names used below. Tool availability and the live input schema take precedence over examples in this skill. Use the live schema for parameter limits and defaults. Do not present example prices in this skill as a current billing quote.


Search for paid Meta ads, organic Instagram posts, and TikTok videos across the Spyglass corpus, or pull a specific account's library with performance sorting and creative metadata.

For a topic, product, or creative format, start with a short phrase in `getMedia` (`query` only), lock the product vertical with `searchTaxonomies` when neighboring categories bleed in, or discover the leading brands with `getAccounts` (`smart` mode) before opening their libraries. For a known brand or creator, resolve the account with `getAccounts`, check `getBrandMediaInsights` or `getBrandCampaigns` for high-level clusters, and pull `getMedia` by `accountId` with no `query`.

## 1. Pick the right mode before calling

`getMedia` runs three different engines behind one tool. Mixing them returns a confident, wrong set.

- **Corpus semantic search (`query` only, no `accountId`)**: Finds creatives across the entire database by visual or spoken similarity.
- **Brand library (`accountId` only, no `query`)**: A direct database table read of one account's ads or posts, sorted by performance or recency.
- **Theme inside a brand (`accountId` + `query`)**: Vector search restricted to one account. Use this only when the user asks how a specific brand advertises a specific product line, ingredient, or angle (e.g., "Primal Queen's iron ads" or "Amazon's pet food ads"). Never add a `query` when the user just wants a brand's general library.

Passing neither `accountId` nor `query` fails validation immediately.

## 2. How to find content on a topic (3-way fallback)

Do not run one `getMedia` call and treat whatever comes back as the whole answer, and do not fire 6 near-identical rephrasings in a loop. Try these three approaches in order when a search comes back thin, off-topic, or dominated by one brand:

1. **Say it the way the ad says it (`query` only)**
   Write a 2–5 word phrase a viewer would see on screen or hear in the hook: `"hydrogel eye patches"`, `"before and after acne"`, `"interactive product quiz"`.
   - If a topic has distinct angles, run 2–3 short, distinct queries (`"peptide serum routine"`, `"copper peptides before and after"`) with `limit: 25`.
   - Leave `sortBy` empty on discovery searches so results stay ranked by vector similarity.

2. **Lock the product category with `searchTaxonomies` (`query` + `filters.taxonomy`)**
   Semantic search bleeds across neighboring categories (`"serum"` pulls hair serum; `"loader"` pulls toy trucks).
   - Call `searchTaxonomies` with the product noun first (free).
   - Pass an exact returned path (e.g., `Beauty > Skincare > Serums`) in `filters.taxonomy` alongside your short `query`.

3. **Find the accounts in the niche, then open their libraries (`getAccounts` → `getBrandMediaInsights` → `getMedia`)**
   Corpus search surfaces scattered posts, not a cohesive campaign. When you need the real leaders in a niche—or when semantic search is noisy—call `getAccounts` in `smart` mode with a concrete market description (`"DTC electrolyte hydration powder brands"`).
   - Pick the 2–3 most relevant accounts.
   - Check `getBrandMediaInsights` (`HOOK` or `AD_ANGLE`) or `getBrandCampaigns` (`CAMPAIGN` or `OFFER`) first to see their creative clusters and active promos.
   - Then pull `getMedia` by `accountId`

Preview of the first 4,000 characters. Read full instructions →