> ## Documentation Index
> Fetch the complete documentation index at: https://docs.terrapinfinance.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Tear Sheets

> Run the same narrative credit write-up across every entity in a watchlist.

[Tear Sheets](https://terrapinfinance.com/platform/tear-sheets) (`/platform/tear-sheets`) is the workspace for narrative write-ups at scale. You write one set of instructions, and the platform produces the same write-up for every entity in a watchlist, each grounded in that entity's own filings.

Use it when you want prose: a credit summary, a memo section, a standard one-pager per name. For figures laid out side by side, use [Data Tables](/platform-guide/data-tables) instead. For a conversation about one credit, use [Deep Dive](/platform-guide/deep-dive/introduction).

## The workspace

There's one workspace **per set of instructions**. The sidebar lists yours under **My tear sheets**, and anyone else's under **Shared with me**.

The main area has the instructions panel on top and a results table below, one row per entity. Each row shows the entity, the date its write-up was generated, and how many source documents went into it. Click a row to expand the write-up inline, with citations linking back to the pages they came from. **Export** saves a single write-up as PDF or DOCX.

Two markers are worth knowing:

* A **New** badge means the write-up arrived since you last looked.
* An **information icon** means the instructions have changed since that write-up was generated, so it no longer reflects what the prompt says today.

## Creating and running a tear sheet

**New Tear Sheet** opens one window covering everything a run needs.

<Steps>
  <Step title="Write the instructions">
    Say what sections you want, in what order, and any definitions the write-up should use. You can load the instructions from one of your existing tear sheets or data tables as a starting point.
  </Step>

  <Step title="Pick the watchlist and scope">
    *Obligor-level* produces one write-up per obligor, pooling every CUSIP under it. *CUSIP-level* produces one per CUSIP, for when bond-level detail matters.
  </Step>

  <Step title="Set the search window and add documents">
    **Years Back** controls how far back filings are searched (1, 2, 3, or 5 years). You can also attach your own uploads to specific entities, and extraction reads them alongside the filings we already found.
  </Step>

  <Step title="Run on a few entities">
    Pick a handful of representative CUSIPs. Extraction covers only your selection, so trying the instructions out is cheap.
  </Step>

  <Step title="Extend to the rest of the watchlist">
    Once the write-ups read the way you want, click **Run** again and select the rest. Entities already covered are locked and keep what they have.
  </Step>
</Steps>

While a run is going, **Cancel** stops it. Completed write-ups are kept; unfinished entities are discarded.

## Staying current

Once a tear sheet covers a watchlist, it keeps up on its own. When new disclosures land for an entity on the list, its write-up is regenerated, and the new one appears with a **New** badge here and in [Surveillance](/platform-guide/surveillance). Adding a CUSIP to the watchlist brings it into the tear sheet too.

## Where they show up

* **Here**, in the workspace, all entities together.
* **On the watchlist**, under [Coverage](/platform-guide/watchlists/coverage).
* **In [Surveillance](/platform-guide/surveillance)**, alongside filings and rating actions, filterable by type.
* **On an obligor page**, under its Tear Sheets tab. You can also start a new one for that obligor from there.

## See also

* [Writing effective prompts](/platform-guide/writing-effective-prompts): structure, scope and definitions that produce comparable write-ups.
* [Data Tables](/platform-guide/data-tables): the same idea for figures rather than prose.