# Decisions

> A decision is a question a person settles. In Fir it is raised while a plan is drafted, always with a recommendation; a panel of other model families gives sealed opinions; the confidence is calibrated on the scope's own decided decisions; people answer in one key or let their own proven rules apply; and every decision joins a history whose effects reach later sessions as In force, while history itself reaches a model only at the level its task needs. Frazer Decisions sells the same panel to any plan or agent. Every command is in the Fir CLI reference: https://frazer.ai/docs/source/reference/cli.md.

Part of Frazer's docs for agents. Every page on frazer.ai: https://frazer.ai/llms.txt. Every docs page: https://frazer.ai/docs/llms.txt.

## How agents should use Decisions

- Raise a question you cannot answer from the code, the records or the decisions in force as a decision row with a recommendation (`fir decision raise`); the panel weighs it.
- Only people answer decisions and decision groups, confirm a conversation's reading and create auto-apply rules; `answer_decision` returns `PERMISSION_DENIED` to agents, and `fir decision talk` with no person present ends with `PERSON_REQUIRED` (exit `13`).
- Use `get_principles` only for your own person's items, to phrase questions and cover options; never show principles to a panel or count them as evidence.
- What a decision binds reaches you as In force. History is precedent, not instruction, and comes only at your task's `lineage` level; ask for one decision at a higher level with `get_decision_lineage`, within your ceiling.
- Never act against a `must`: raise a decision row that proposes to supersede it, with new evidence.

## Pages

- [Decisions](https://frazer.ai/docs/source/decisions.md?from=llms-txt): Decisions inside plans in Fir: raised while a plan is drafted, always with a recommendation, weighed by a panel of other model families with sealed opinions, shown with a calibrated confidence on a page that reads verdict, opinions, evidence and consequences, answered by people in one key or applied by their own proven rules, grouped at scale, kept in a decision history that models see only at the level their task needs, and sold to any plan or agent as Frazer Decisions.
- [Decide by conversation](https://frazer.ai/docs/source/decisions/conversation.md?from=llms-txt): Answer your queue of decisions and approvals in a conversation, by voice, dictation or text: one question at a time with its recommendation, a read-back of what the agent understood that writes nothing until you confirm, a queue plan for a backlog of 200, keep mine while away, and principles learned from your own decided answers, cited, kept by you and never shown to the panel.

## Optional

- [Full text of these pages](https://frazer.ai/docs/source/decisions/llms-full.txt): 2 pages, 71,696 characters, about 17,924 tokens