@softspark/jira-mcp 1.15.0 → 1.16.1

package/CHANGELOG.md

@@ -7,6 +7,79 @@ Versioning follows [Semantic Versioning](https://semver.org/).
 
 ---
 
+## v1.16.1 -- Bad arguments stop answering UNKNOWN_ERROR (2026-09-18)
+
+Both packages change in this release.
+
+### Fixed
+
+- **Argument checks answer `INVALID_INPUT`.** Every input check in both servers
+  raised a bare `Error`, and a tool response takes its `code` only from the
+  typed hierarchy, so "you passed both `content` and `storage`" came back as
+  `UNKNOWN_ERROR`. That code reads as "something broke", and an agent that sees
+  it stops or looks for a way around the server instead of fixing its call. 31
+  sites now raise `InvalidInputError`: a missing required argument, two
+  arguments that exclude each other, an update that names nothing to change, a
+  malformed time (`2d`) or Tempo date, an unknown `group_by`, category or
+  `position` value, a file over the upload limit, an unknown tool name. Nothing
+  reaches Atlassian when it is raised, and the messages are unchanged.
+- **`TEMPLATE_MISSING_VAR` is real now.** `create_task`,
+  `add_templated_comment` and `create_page` answer it when a template is
+  rendered without a required variable. The class and its row in the API
+  reference had existed since templates shipped, but nothing raised it.
+- **`add_page_inline_comment` answers `ANCHOR_NOT_FOUND`** when
+  `text_selection` is not in the page body, instead of `UNKNOWN_ERROR`. Found
+  in the 1.16.0 post-release run.
+- A raw NUL byte in a string literal in `tempo-operations.ts` made git and grep
+  treat the file as binary: its diffs showed as `Bin` and searches skipped it
+  without saying so. It is now the escape `'\u0000'` behind a named constant.
+  Behaviour is identical.
+- The README test count had been stale since 1.16.0.
+
+### Added
+
+- `packages/core/tests/typed-errors.test.ts` fails the build on a bare
+  `new Error(` anywhere a tool call can reach (`src/cli/` is exempt, a CLI error
+  goes to a terminal) and on a raw control byte in a source file. It found six
+  of the 31 sites on its first run, in the file grep could not read.
+
+### Upgrade note
+
+A client that matched on `UNKNOWN_ERROR` to detect a rejected call now sees
+`INVALID_INPUT`, `TEMPLATE_MISSING_VAR` or `ANCHOR_NOT_FOUND` there.
+`UNKNOWN_ERROR` is left for failures nobody anticipated.
+
+## v1.16.0 -- Creator and reporter on every task read (2026-09-18)
+
+`@softspark/confluence-mcp` has no behaviour change in this release; it is bumped
+together with `@softspark/jira-mcp` per ADR-0002.
+
+### Added
+
+- **`creator` and `reporter` in task reads.** `search_tasks`,
+  `get_task_details`, `sync_tasks` and `read_cached_tasks` now return who filed
+  an issue and who it is reported by. Until now the connector fetched `creator`
+  only for the `delete_task` ownership guard and never asked Jira for
+  `reporter`, so an agent asked "who raised this bug" had nothing to answer
+  from and went looking for a way around the server. Both fields hold the
+  email, or the display name when Atlassian privacy settings hide the email,
+  and `null` when Jira returns none. `reporter` is editable in Jira and can be
+  empty; `creator` never changes.
+- The `search_tasks` and `get_task_details` tool descriptions name the fields a
+  result carries, so a client can tell what is available without a trial call.
+
+### Changed
+
+- **Task cache rows carry `creator` and `reporter`.** The schema defaults both
+  to `null`, so a cache file written by an older version still loads and
+  `CACHE_VERSION` stays `1.0`. Rows cached before the upgrade read `null` until
+  the next `sync_tasks`.
+
+### Fixed
+
+- `kb/reference/api.md` showed `get_task_details` answering in a plain-text
+  layout the tool never produced. The example is now the JSON it returns.
+
 ## v1.15.0 -- Tempo worklogs and reports (2026-09-17)
 
 `@softspark/confluence-mcp` has no behaviour change in this release; it is bumped

package/README.md

@@ -4,17 +4,19 @@
 
 [![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-1.15.0-blue)](CHANGELOG.md)
+[![version](https://img.shields.io/badge/version-1.16.1-blue)](CHANGELOG.md)
 [![License: Apache 2.0](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](LICENSE)
 
 ---
 
-## What's New in v1.15.0
+## What's New in v1.16.1
 
-- **Tempo worklogs and reports.** `search_tempo_worklogs` lists Tempo Timesheets entries in a date range with issue keys and user names resolved; `get_tempo_report` sums hours by project, user and/or task. Filters combine, so "this person's hours on this project last month" is one call.
-- Tempo REST API v4 answers in numeric ids only, so every result is joined against Jira. An issue the token cannot browse keeps its hours under `#<id>` instead of vanishing from a total.
-- `jira-mcp config set-tempo-token` stores the Tempo token next to the Jira credential, per site with `--url`. Nothing else needs it: without a token the two Tempo tools answer `TEMPO_NOT_CONFIGURED` and every other tool works as before.
-- Fixed: `config set-credentials` no longer drops a stored Tempo token when the Jira token is rotated.
+- **A rejected call says so.** Every argument check used to answer `UNKNOWN_ERROR`, which reads as "something broke". They now answer `INVALID_INPUT`: a missing argument, `template_id` together with `markdown`, an update with nothing to change, a time like `2d`, a malformed Tempo date or `group_by`. Nothing reaches Jira, and the message names what to pass.
+- A template rendered without a required variable answers `TEMPLATE_MISSING_VAR`, as the API reference always said it did.
+- If a client matched on `UNKNOWN_ERROR` to detect a bad call, match on the new codes instead. `UNKNOWN_ERROR` is left for failures nobody anticipated.
+- **From 1.16.0: `creator` and `reporter` on every task read.** `search_tasks`, `get_task_details`, `sync_tasks` and `read_cached_tasks` now say who filed an issue and who it is reported by. Before, neither field reached a response, so "who raised this bug" had no answer.
+- Both hold the email, or the display name when Atlassian privacy settings hide the email, and `null` when Jira returns none. To filter by them, use JQL: `creator = "pm@example.com"` or `reporter = currentUser()`.
+- An existing task cache keeps loading. Rows cached before the upgrade read `null` for both fields until the next `sync_tasks`.
 - Released together with `@softspark/confluence-mcp` under one version. See the [changelog](CHANGELOG.md).
 
 ## Table of Contents
@@ -207,10 +209,11 @@ Configuration is automatically loaded from `~/.softspark/jira-mcp/`. Run `jira-m
 If you use `@softspark/ai-toolkit`, register the Jira rules so that Claude Code automatically follows project conventions (language checks, sync-first workflow, time format, etc.):
 
 ```bash
-ai-toolkit rules add jira-mcp --path /path/to/jira-mcp/rules/jira-mcp.md
+ai-toolkit add-rule https://raw.githubusercontent.com/softspark/jira-mcp/main/rules/jira-mcp.md
+ai-toolkit update
 ```
 
-Or copy `rules/jira-mcp.md` to your ai-toolkit rules directory manually. The rules file covers:
+The URL form is re-fetched on every `ai-toolkit update`, so the rules follow releases. A local path works too (`ai-toolkit add-rule /path/to/jira-mcp/rules/jira-mcp.md`) but stays frozen at that copy. The rules file covers:
 - **Language first** -- always check project language before writing comments/descriptions
 - **Sync before read** -- cache may be stale
 - **Status transitions** -- check valid transitions before changing status
@@ -288,9 +291,9 @@ src/
 
 **Supply chain protection** -- `ignore-scripts=true`, no axios, no dynamic requires. Self-contained 561KB library bundle, 1 runtime dep (commander).
 
-**Typed error hierarchy** -- 30 error classes with machine-readable codes. Every tool returns structured `{ success, error, code }` responses. No stack traces leak to MCP clients.
+**Typed error hierarchy** -- 32 error classes with machine-readable codes. Every tool returns structured `{ success, error, code }` responses. No stack traces leak to MCP clients.
 
-**Strict TypeScript** -- `strict: true`, no `any`, `readonly` interfaces, Zod validation at all boundaries, 1093 tests across 85 test files.
+**Strict TypeScript** -- `strict: true`, no `any`, `readonly` interfaces, Zod validation at all boundaries, 1109 tests across 86 test files.
 
 ## Documentation