# The Book: where your production stands, its plan and its lessons

> In short: The Book tab shows what CODA knows about your production: where it stands, a plan you and CODA keep, lessons CODA follows, and an activity feed of everything that happened. CODA reads the same pages you do, and a long conversation no longer loses what was settled early.

Applies to: v0.14 · Last reviewed: 2026-11-02 · Canonical: https://docs.zymcos.com/product/the-production-book/

The Book is a tab in the right column. It holds everything CODA knows about a production, in
seven chapters, so nothing she relies on is hidden from you. CODA reads exactly what you read
here.

## The seven chapters

- **Situation.** Where the production stands right now: how many scenes, which have shots, which
  cards are waiting, which takes failed, and the areas that need a look. It is worked out by the
  app from the production itself, never typed by anyone, and it updates a few seconds after
  something changes. Ask CODA "where are we?" and she answers from it.
- **Plan.** What is next. Each item is *proposed*, *next*, *doing*, *done* or *dropped*. Add one
  in the box ("Add to the plan"), or tell CODA: "add 'reshoot the door' to the plan." CODA adds
  an item at once when you ask in your own words; what she suggests on her own she leaves as
  *proposed*. She does not mark an item *done* just because she says so; the app checks it
  against the production, and an item that is not true yet stays open.
- **Lessons.** Things CODA should remember to do, or not do, in this production, such as "stills
  before any video" or "keep prompts short". Say "from now on…" and ask her to remember it, and
  CODA proposes a lesson for you to accept. In the Book you can accept a lesson, correct it or retire it, and keep it for one production or
  for your whole workspace. A failed take also leaves a lesson, written by the app,
  so the same mistake is not repeated. A lesson for the whole workspace cannot name one production's entries or a person.
- **Decisions** and **Charter.** What was decided and the production's rules (see *Decisions and
  the charter*).
- **Activity.** Everything that happened, newest first, each as one sentence, with who did it (you,
  CODA for you, a collaborator, a reviewer or the app), without names. Above it, a short summary by
  day. A new conversation opens with what changed since you were last there. See
  [Activity](https://docs.zymcos.com/product/activity/).
- **Playbook.** The playbook the production is following, if any.

## CODA can search the Book

Ask "what did we decide about the lighthouse?" and CODA looks through every chapter and names the
rows she found. If she cites a row that does not exist, or one that was closed, the app says so.

## Long conversations

When a conversation grows past what CODA can read at once, the app keeps a short record of it,
made from your messages, the cards and what CODA did, in view in place of the older messages. A
colour you chose forty messages ago, or a name you gave a character, is still held, and a value you
later changed is the changed one. You can switch this off in **Settings**, under CODA, with
*Record a long conversation for CODA as it goes*.

## How much CODA reads each turn

Four limits in **Settings** (CODA) set how many open items, plan items, lessons and scenes CODA
reads in a turn. Your plan sets sensible defaults and you can lower them. A longer script is a
window of scenes around the one you have open, so a feature film costs what a commercial does.

## Related

- *Asking CODA to do something to a take*: the cards CODA puts up and how she reports on them.
- *Settings*: where the limits live.
