Skip to content

Shortcut

REST-only (no GraphQL), so structurally closer to Jira than to the GraphQL trackers. Selection is a raw Shortcut search string — the same syntax the in-app Search Stories box accepts.

1. API token

app.shortcut.com/settings/account/api-tokens → Generate Token → copy the value.

The Read-only checkbox controls whether the token can run mutations. Leave it off if you plan to use mutations:, on otherwise.

Format: sct_rw_<workspace>_<random> (read-write) or sct_ro_<workspace>_<random> (read-only). Sent as Shortcut-Token: <token> — a custom header, not Authorization: Bearer.

2. Configure

Shortcut:
  apiToken: "sct_rw_workspace_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"

The token's workspace is implicit — Preesta calls GET /api/v3/member once at startup and reads workspace2.url_slug to build app.shortcut.com/<slug>/… deep links.

3. Write rules

rules:
  - tracker: shortcut
    tags: morning
    filter: "state:\"In Progress\" type:bug !is:archived"
    notify:
      subject: "Open Shortcut bugs"
      mailTo: assignee

  - tracker: shortcut
    filter: "team:platform !state:completed has:deadline deadline:<today"
    notify:
      subject: "Overdue Shortcut stories"
      mailTo: reporter

filter: is Shortcut's own search-operator syntax — state:"...", type:, label:, owner:, requester:, has:, is:, deadline:, team:, with ! to negate. The web UI uses exactly the same syntax in its top search bar.

No identity in filters. owner:me / requester:me resolve to the token's owner, which is almost never who the digest is for. Use mailTo markers for per-recipient fan-out.

Foreign-ID resolution

Shortcut's search response gives you workflow state IDs (integers) and owner IDs (UUIDs), not names or emails. Preesta resolves them via two auxiliary REST calls cached per process:

  • GET /api/v3/workflows → workflow state ID → human-readable state name ("In Progress", "Done", …)
  • GET /api/v3/members → owner UUID → { name, email }

Both are fetched lazily on the first GetIssues call. Failures (HTTP error, restricted token) log a warning and degrade gracefully — owner email becomes empty, marker routing skips that recipient, the digest still goes to others.

What appears in a Shortcut digest

Each item shows the key (sc-26 — Shortcut's own branch-naming convention), the title, and any columns: you ask for. Shortcut-specific notes:

  • Type is the story type (feature / bug / chore).
  • Status is the workflow state name (e.g. In Progress, Done).
  • The story's requester fills both Reporter and Creator columns.
  • Resolution isn't separately exposed — the workflow state name carries that meaning (Completed, Cancelled).
  • DueDate comes from the story's deadline.

Mutations

REST verb / urlPattern / body shape, same as Jira:

- tracker: shortcut
  filter: "state:\"Ready for Review\" updated:<-7d"
  mutations:
    - verb: POST
      urlPattern: "https://api.app.shortcut.com/api/v3/stories/{{@issueId}}/comments"
      body: |
        { "text": "Stale — please update or close." }
    - verb: PUT
      urlPattern: "https://api.app.shortcut.com/api/v3/stories/{{@issueId}}"
      body: |
        { "archived": true }

{{@issueId}} is the numeric story ID. Requires a sct_rw_* (write-enabled) token. Per-mutation failures are logged + skipped.

The digest header link points at https://app.shortcut.com/<workspace>/search#<filter> — Shortcut uses a URL fragment (after #) for search queries, not a query parameter. Fragment encoding is minimal (only # and whitespace escaped) so the filter renders unmodified in the Search Stories box. Verified live in Chrome.