Skip to content

scry v0.3.0

Choose a tag to compare

View all tags
@github-actions github-actions released this 05 Aug 08:38
v0.3.0
5585567

The mirror configures itself

Until now, custom fields required a hand-written fieldMap — so in practice
nobody configured them and every board looked half-empty. v0.3.0 removes the
step entirely:

  • Custom-field auto-discovery. The first full sync fetches every field,
    groups the site catalog by display name (Jira mints a separate field id per
    board template for the same concept — on one large site, 57 of 353 names
    carried two or more ids), classifies each field's role and editor from its
    schema, and backfills the mirror from the already-stored raw documents with
    no re-download. The sync log tells you what it did:

    fields: discovered 23 custom fields in use — labels, filters, and editors configured
fields: backfilled 6,543 issues from the mirror (no re-download)

  • Projects are optional. scry init needs only site, email, and token.
    An empty project list means every project the account can see; the sync
    start line shows the total so you can narrow later.

  • The UI follows the data. Detail panels show only the fields an issue
    actually fills, with Jira's own display names; long-form fields (repro
    steps, QA notes) render as document sections, not chips. Filter axes come
    from the discovered fields per project scope — a team board is not offered
    a QA board's thirty fields. array<option> fields get a proper
    multi-select editor, values from Jira's own allowed list.

  • Everything stays editable: Settings → Fields in the web UI (edits are
    pinned across re-discovery), scry fields --apply, or the fields array
    in ~/.scry/config.json.

Also in this release

  • Readable sync progress — 1,200 / 6,543 (18%) with an upfront count
    and elapsed time, instead of a page counter with no denominator.
  • Sync history — click the sidebar timestamp for recent runs (kind,
    changed/deleted counts, errors). Meaningful runs only; no-op polls are not
    recorded.
  • Agent-driven onboarding — docs/AGENT_SETUP.md opens with a paste-ready
    prompt that lets a coding agent move an existing Jira setup (env vars, MCP
    connector) into scry end to end.
  • The group-taxonomy field is renamed to the neutral team_group.

Install / upgrade

brew install midagedev/tap/scry     # or: brew upgrade midagedev/tap/scry
# or
curl -fsSL https://raw.githubusercontent.com/midagedev/scry/main/scripts/install.sh | sh

macOS binaries are signed and notarized. Existing mirrors: run
scry sync --full once after upgrading to trigger field discovery.

Assets 9
Loading