Init Jira

---
name: init-jira
description: >
  Set up the Jira Cloud integration for this project. Verifies credentials
  against a real Jira Cloud tenant, discovers the tenant's custom-field
  catalogue and project list, and persists the results under
  `<paths.integrations>/jira/` (default `.accelerator/state/integrations/jira/`)
  as team-shared, version-controlled JSON caches. Idempotent: safe to re-run
  after credential or project changes.
argument-hint: "[--site <subdomain>] [--email <addr>] [--refresh-fields] [--list-projects] [--list-fields]"
disable-model-invocation: true
allowed-tools:
  - Bash(accelerator config *)
  - Bash(accelerator jira *)
---

# Init Jira

!`accelerator config context --skill init-jira --fail-safe`

> **Configuration**: Set `work.integration: jira` and
> `jira.project_key: <KEY>` in `.accelerator/config.md` to
> enable auto-scoping. See the
> [`### work` section of `configure/SKILL.md`](../../config/configure/SKILL.md#work)
> for the full reference.

You are setting up the Jira Cloud integration for this project. Work through
the steps below in order, stopping to prompt the user only when a value is
missing and cannot be derived from existing configuration.

Run every `accelerator jira …` and `accelerator config …` invocation **directly**
as an executable; never prefix it with `bash`/`sh`/`env` and never pipe its
output (a wrapper prefix or a pipe escapes the skill's `allowed-tools`
permission and forces an unnecessary prompt).

## Step 0: Parse arguments

Read the argument string (if any) and note:

- `--site <subdomain>` — Jira Cloud subdomain override (e.g. `atomic-innovation`)
- `--email <addr>` — Atlassian account email override
- `--refresh-fields` — re-run field discovery only
- `--list-projects` — print cached projects and exit (no network call)
- `--list-fields` — print cached fields and exit (no network call)

If `--list-projects` was requested, run:

```
accelerator jira init list-projects
```

If `--list-fields` was requested, run:

```
accelerator jira init list-fields
```

Both print the cached JSON array; render it as a readable list. If
`--refresh-fields` was requested, skip to Step 5 (field discovery only).

## Step 1: Resolve site

Use the site from `--site` if provided. Otherwise read it from config:
`accelerator config get jira.site`. If still empty,
prompt: *"Enter your Jira Cloud subdomain (the part before `.atlassian.net`,
e.g. `mycompany`):"*

## Step 2: Resolve email

Use `--email` if provided. Otherwise read it from config:
`accelerator config get jira.email`. If still empty,
prompt: *"Enter your Atlassian account email:"*

## Step 3: Verify and persist site.json

Run:

```
accelerator jira init verify
```

Credential resolution folds into this one call; the token is never printed. On
success the subcommand emits a JSON document with `outcome: "verified"` plus
`{site, accountId}`, and writes
`.accelerator/state/integrations/jira/site.json`. Print: *"Verified as
`<accountId>` on `<site>.atlassian.net`."*

If it fails naming a missing token, tell the user:

> No Jira API token found. Generate one at
> <https://id.atlassian.com/manage-profile/security/api-tokens>, then add it
> to `.accelerator/config.local.md` (which is gitignored):
>
> ```yaml
> ---
> jira:
>   token_cmd: "op read op://Work/Atlassian/credential"
> ---
> ```
>
> Re-run `/init-jira` once the token is configured.

Then stop. On any other non-zero exit, show the error and stop.

## Step 4: Discover projects and fields

Run:

```
accelerator jira init discover
```

On `outcome: "discovered"` the subcommand writes `projects.json` (project
key/id/name) and `fields.json` (field id/key/name/slug/schema) atomically. Both
are byte-idempotent — re-running against an unchanged tenant produces no diff.

If `--refresh-fields` was requested, run only:

```
accelerator jira init refresh-fields
```

## Step 5: Default project key

Run:

```
accelerator jira init prompt-default
```

On `default-prompted` the subcommand reports the resolved default project. If
`jira.project_key` is already set this is a no-op; otherwise offer to
write the chosen key to `config.md`.

## Step 6: Confirm completion

Print a summary:

```
Jira integration initialised:
  Site:     <site>.atlassian.net
  Fields:   <N> fields cached (.accelerator/state/integrations/jira/fields.json)
  Projects: <M> projects cached (.accelerator/state/integrations/jira/projects.json)
  Default:  <KEY> (jira.project_key)
```

Remind the user to commit `.accelerator/state/integrations/jira/{fields,projects}.json`
so teammates pick up the shared cache without running `/init-jira` themselves.
(`site.json` is gitignored — each developer runs `/init-jira` to configure their
own credentials.)

!`accelerator config instructions init-jira --fail-safe`