@softspark/jira-mcp 1.7.0 → 1.8.1

package/CHANGELOG.md

@@ -7,7 +7,46 @@ Versioning follows [Semantic Versioning](https://semver.org/).
 
 ---
 
-## v1.7.0 -- Apache-2.0 (2026-07-27)
+## v1.8.1 -- Duplicate Detection (2026-08-03)
+
+### Fixed
+
+- **Duplicate detection never matched anything.** `findExistingTask` searched with
+  `summary = "..."`, but `summary` is a text field and JQL only supports `~` on it.
+  Jira answers `=` with an empty result set instead of an error, so every lookup
+  reported "not found": `update_existing` never updated, and re-running a bulk config
+  created a second copy of every task. The lookup now uses a quoted `~` phrase and
+  compares the returned summary exactly, since `~` is a fuzzy match. Found by running
+  a config twice against a live instance and getting a duplicate issue.
+
+## v1.8.0 -- Status Paths (2026-08-03)
+
+### Added
+
+- **`status` accepts an ordered path** in bulk task configs -- `"status": ["On hold", "Open"]`
+  walks the transitions one at a time. Jira only exposes the transitions available from an
+  issue's *current* status, so a target that is not directly reachable from the initial
+  status could not be set at all before. A plain string still means a single transition.
+- **`warnings` in `create_monthly_tasks` output** -- per-task problems that did not stop the
+  issue from being written now surface in the tool response instead of being dropped by the
+  counters-only summary.
+
+### Fixed
+
+- **A rejected status transition is no longer silent.** `setStatus` swallowed every failure,
+  including the ordinary case of the requested status simply not being reachable: the issue
+  was created, the status was ignored, and the run reported `failed: 0`. Eleven monthly admin
+  tasks sat in the wrong status for months because of this. The transition is now reported as
+  a `warning` on the task result, naming the status that failed and listing the ones that were
+  reachable at that point. The issue itself is still created -- the transition stays non-fatal.
+
+### Changed
+
+- **`TaskResult` gains a `warning` field** (`string | null`). `formatBulkResult` prints it
+  indented under the task line. Consumers destructuring `TaskResult` are unaffected; anyone
+  constructing one now has to supply the field.
+
+
 
 ### Changed -- licence: MIT to Apache-2.0
 

package/README.md

@@ -4,17 +4,17 @@
 
 [![CI](https://github.com/softspark/jira-mcp/actions/workflows/ci.yml/badge.svg)](https://github.com/softspark/jira-mcp/actions/workflows/ci.yml)
 [![npm](https://img.shields.io/npm/v/@softspark/jira-mcp)](https://www.npmjs.com/package/@softspark/jira-mcp)
-[![version](https://img.shields.io/badge/version-1.7.0-blue)](CHANGELOG.md)
+[![version](https://img.shields.io/badge/version-1.8.1-blue)](CHANGELOG.md)
 [![License: Apache 2.0](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](LICENSE)
 
 ---
 
-## What's New in v1.7.0
+## What's New in v1.8.1
 
-- **Licence: MIT to Apache-2.0.** Still permissive -- fork it, modify it, ship it commercially. What changes is what a redistributor owes back: the contents of [NOTICE](NOTICE) must travel with any redistribution (§4d), modified files must say they were modified (§4b), plus an express patent grant and a trademark reservation. Releases up to v1.6.0 stay available under MIT; nothing already granted is revoked.
-- **Attribution reaches the published bundles.** npm ships `dist/`, not `src/`, and the build minifies away every comment -- so the SPDX headers added to 165 source files never reach a consumer. A build banner carries the licence into `dist/index.js` and `dist/cli.js` instead, and a test asserts both entries keep it.
-- **Licensing gate in `npm test`** -- `tests/licensing.test.ts`, 8 assertions covering headers, `LICENSE`, `NOTICE`, npm `files`, and the build banner. Release SOP gains Phase 4.7 running the same gate before tagging.
-- **Release SOP hardened** -- tag placement is now asserted before the push, and `git push --tags` is replaced with a single-tag push by full ref. A `--tags` push is how a sibling project silently skipped a publish.
+- **Duplicate detection actually works now.** The lookup used `summary = "..."`, and Jira answers `=` on a text field with an empty result set rather than an error -- so it always reported "not found". `update_existing` never updated anything, and re-running a bulk config duplicated every task. Now a quoted `~` phrase with an exact comparison on the result.
+- **`status` takes a path, not just a name.** `"status": ["On hold", "Open"]` walks the transitions in order. Jira only ever exposes the transitions out of an issue's *current* status, so a target that is not directly reachable from the initial status was previously impossible to set -- which is exactly how workflows that gate `Open` behind a `Reopen` transition behave.
+- **Failed transitions stop being silent.** The old code caught every transition error and moved on: issue created, status ignored, run reports `failed: 0`. It now returns a `warning` naming the status that failed and listing what was reachable instead, surfaced per task and aggregated into `create_monthly_tasks` output.
+- **`TaskResult.warning`** -- new `string | null` field for problems that occur after the issue is already written. `formatBulkResult` prints it under the task line.
 
 See [CHANGELOG.md](CHANGELOG.md) for full details.