> ## Documentation Index
> Fetch the complete documentation index at: https://docs.no-tickets.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Frontmatter

> Exhaustive field reference for every YAML frontmatter field on epics, features, and fixes.

YAML frontmatter is the structured slice of every no-tickets file. The
parser reads frontmatter strictly; unknown top-level fields are rejected.
Tool-specific data goes under `meta:` (see below).

The frontmatter block must be the first content in the file — three dashes,
the YAML body, three dashes, then the markdown.

```markdown theme={null}
---
id: my-feature
type: feature
...
---

# My feature
```

## Field reference

### Base fields (all document types)

| Field     | Type                                          | Required | Notes                                                                                                      |
| --------- | --------------------------------------------- | :------: | ---------------------------------------------------------------------------------------------------------- |
| `id`      | kebab-case string                             |     ✓    | Must match the filename (without `.md`), or the directory name for `epic.md`. Lowercase, hyphen-separated. |
| `type`    | `epic` \| `feature` \| `fix`                  |     ✓    | Document type — determines which other fields are required.                                                |
| `title`   | string                                        |     ✓    | Human-readable. Shown on the board, in the feed, in search.                                                |
| `status`  | `not_started` \| `in_progress` \| `completed` |     ✓    | Workflow status — orthogonal to `phase`.                                                                   |
| `created` | `YYYY-MM-DD`                                  |     ✓    | Creation date. Set once and never updated.                                                                 |
| `updated` | `YYYY-MM-DD`                                  |     ✓    | Last meaningful update. The CLI bumps this on save.                                                        |
| `meta`    | object                                        |     —    | Extensible bag for tool-specific data. See below.                                                          |

### Feature / fix fields

| Field           | Type                                                           | Required | Notes                                                                   |
| --------------- | -------------------------------------------------------------- | :------: | ----------------------------------------------------------------------- |
| `epic`          | kebab-case string                                              |     ✓    | Parent epic ID. Must match an existing `.tiny-brain/<epic>/` directory. |
| `phase`         | `ideation` \| `development` \| `testing` \| `review` \| `done` |     ✓    | Lifecycle phase — see [Phases](/concepts/phases).                       |
| `assignee`      | string                                                         |     —    | Member name or agent ID. Empty/missing means unassigned.                |
| `assignee_type` | `human` \| `agent`                                             |     —    | Required if `assignee` is set. Drives avatar + filter.                  |

### Fix-only fields

| Field      | Type                                      | Required | Notes            |
| ---------- | ----------------------------------------- | :------: | ---------------- |
| `severity` | `critical` \| `high` \| `medium` \| `low` |     ✓    | Triage priority. |

## `meta` — the escape hatch

`meta:` is the one field tools can extend without breaking the parser.
Anything under `meta:` is preserved on save and round-trips through pushes
unchanged.

Conventions:

* **Namespace your keys.** Tools should write under a sub-object named after
  the tool: `meta.tiny_brain.quality_score`, not `meta.quality_score`.
* **Don't put core data here.** If a field is part of the spec, it belongs
  at the top level, not under `meta`.

Worked example:

```yaml theme={null}
meta:
  tiny_brain:
    quality_score: 82
    tdd_compliant: true
  internal_tracker:
    jira_key: AUTH-1234
```

## Field validation

The parser enforces a few rules beyond "must be the right type":

1. **`id` must match the filename.** A file named `google-oauth.md` with
   `id: googleoauth` is rejected on push.
2. **Dates must be `YYYY-MM-DD`.** No timezones, no times.
3. **`epic` must reference an existing epic.** A feature referencing a
   non-existent epic is rejected on push.
4. **`assignee_type` is required when `assignee` is set.** Either both, or
   neither.

## Worked examples

### Minimal epic

```yaml theme={null}
---
id: auth-flow
type: epic
title: Authentication Flow
status: in_progress
created: 2026-05-20
updated: 2026-05-21
---
```

### Feature assigned to an agent

```yaml theme={null}
---
id: google-oauth
type: feature
epic: auth-flow
title: Google OAuth sign-in
phase: development
status: in_progress
assignee: claude-coder
assignee_type: agent
created: 2026-05-20
updated: 2026-05-21
meta:
  tiny_brain:
    quality_score: 78
---
```

### Critical fix

```yaml theme={null}
---
id: logout-redirect-bug
type: fix
epic: auth-flow
title: Logout redirects to /undefined
severity: critical
phase: testing
status: in_progress
assignee: maria
assignee_type: human
created: 2026-05-20
updated: 2026-05-21
---
```

## See also

* [Markdown format → Epics](/markdown-format/epics) / [Features](/markdown-format/features) / [Fixes](/markdown-format/fixes)
* [Concepts → Phases](/concepts/phases) — what each `phase` value means
* [Concepts → Roles](/concepts/roles) — `assignee_type` semantics