norn-cli 1.10.6 → 1.12.0

package/.kanbn/index.md

@@ -0,0 +1,30 @@
+---
+startedColumns:
+  - 'In Progress'
+completedColumns:
+  - Done
+---
+
+# Norn
+
+## Backlog
+
+- [do-5-customer-conversations](tasks/do-5-customer-conversations.md)
+- [write-down-repeated-wording](tasks/write-down-repeated-wording.md)
+- [refine-your-pitch](tasks/refine-your-pitch.md)
+- [recruit-2–3-pilot-users](tasks/recruit-2–3-pilot-users.md)
+- [decide-what-success-in-a-pilot-looks-like](tasks/decide-what-success-in-a-pilot-looks-like.md)
+- [interview-script](tasks/interview-script.md)
+
+## Todo
+
+- [finalise-the-one-line-pitch](tasks/finalise-the-one-line-pitch.md)
+- [make-a-list-of-10-people-to-speak-to](tasks/make-a-list-of-10-people-to-speak-to.md)
+- [prepare-your-customer-interview-questions](tasks/prepare-your-customer-interview-questions.md)
+
+## In Progress
+
+## Done
+
+- [book-first-mentor-session](tasks/book-first-mentor-session.md)
+- [write-the-one-pager](tasks/write-the-one-pager.md)

package/.kanbn/tasks/book-first-mentor-session.md

@@ -0,0 +1,13 @@
+---
+created: 2026-04-09T22:55:34.350Z
+updated: 2026-04-12T12:19:50.345Z
+assigned: ""
+progress: 0
+tags:
+  - 'Week One'
+due: 2026-04-12T00:00:00.000Z
+started: 2026-04-12T09:08:33.269Z
+completed: 2026-04-12T12:19:50.345Z
+---
+
+# Book first mentor session

package/.kanbn/tasks/decide-what-success-in-a-pilot-looks-like.md

@@ -0,0 +1,9 @@
+---
+created: 2026-04-09T23:04:43.339Z
+updated: 2026-04-09T23:04:43.337Z
+assigned: ""
+progress: 0
+tags: []
+---
+
+# Decide what success in a pilot looks like

package/.kanbn/tasks/do-5-customer-conversations.md

@@ -0,0 +1,9 @@
+---
+created: 2026-04-09T23:03:31.570Z
+updated: 2026-04-09T23:03:31.569Z
+assigned: ""
+progress: 0
+tags: []
+---
+
+# Do 5 customer conversations

package/.kanbn/tasks/finalise-the-one-line-pitch.md

@@ -0,0 +1,11 @@
+---
+created: 2026-04-09T22:59:46.277Z
+updated: 2026-04-09T23:03:00.808Z
+assigned: ""
+progress: 0
+tags:
+  - 'Week One'
+due: 2026-04-12T00:00:00.000Z
+---
+
+# Finalise the one-line pitch

package/.kanbn/tasks/interview-script.md

@@ -0,0 +1,49 @@
+---
+created: 2026-04-09T23:25:42.568Z
+updated: 2026-04-09T23:25:42.565Z
+assigned: ""
+progress: 0
+tags: []
+---
+
+# Interview Script
+
+This is the important bit. You are not trying to impress them. You are trying to learn where the pain is.
+Opening
+“Thanks for taking the time. I’m working on a VS Code-native API testing and automation tool called Norn. I’m speaking to people to understand how teams currently handle API testing, bug reproduction, and automation. I’m not here to hard-sell you anything — I mainly want to learn how you do things today and where the pain is.”
+Warm-up
+“Can you tell me a bit about your role and how involved you are with API testing or debugging?”
+Current workflow
+“How do you currently test APIs day to day?”
+“What tools do you use for exploratory API work?”
+“What tools do you use for repeatable automated API tests?”
+“Are those the same tools, or different?”
+“Who usually owns API regression coverage in your team?”
+“When there’s a bug, how do developers usually reproduce it?”
+Pain discovery
+“What’s the most annoying part of your current setup?”
+“Where does the process break down?”
+“Do developers and QA use the same tools or mostly different ones?”
+“Have you ever had useful API tests or requests trapped in one place where other people don’t really use them?”
+“What becomes hard to maintain as the project grows?”
+“If you could magically remove one frustration from your API workflow, what would it be?”
+Team and buying context
+“If a new tool genuinely improved this, who would care most?”
+“Who would likely use it first?”
+“Who would need to approve it?”
+“Would this be more of a team decision, an engineering decision, or something individuals would adopt first?”
+Reaction to the concept
+After you’ve listened first, say this:
+“I’m building something that lets developers and QA write readable requests, reusable definitions, environments, assertions, and multi-step sequences directly in VS Code, with the aim of reducing the gap between exploratory work and automation. Based on what you’ve said, what parts of that sound useful and what parts don’t?”
+Then ask:
+“What would make you want to try something like that?”
+“What would make you ignore it?”
+“What would you compare it against immediately?”
+“What would it need to do well before you’d take it seriously?”
+Pilot questions
+“If I gave you access to try this on a real workflow, what would be the best use case to test first?”
+“What would success look like for you in a pilot?”
+“What would need to happen for you to keep using it after the trial?”
+“If it worked well, would you see this as something for just you, or for your wider team?”
+Close
+“This has been really useful. Would you be open to a follow-up once I’ve tightened the product and pilot setup?”

package/.kanbn/tasks/make-a-list-of-10-people-to-speak-to.md

@@ -0,0 +1,11 @@
+---
+created: 2026-04-09T23:00:11.428Z
+updated: 2026-04-09T23:02:58.760Z
+assigned: ""
+progress: 0
+tags:
+  - 'Week One'
+due: 2026-04-12T00:00:00.000Z
+---
+
+# Make a list of 10 people to speak to

package/.kanbn/tasks/prepare-your-customer-interview-questions.md

@@ -0,0 +1,11 @@
+---
+created: 2026-04-09T23:00:42.150Z
+updated: 2026-04-09T23:02:55.974Z
+assigned: ""
+progress: 0
+tags:
+  - 'Week One'
+due: 2026-04-12T00:00:00.000Z
+---
+
+# Prepare your customer interview questions

package/.kanbn/tasks/recruit-2/342/200/2233-pilot-users.md

@@ -0,0 +1,9 @@
+---
+created: 2026-04-09T23:04:26.660Z
+updated: 2026-04-09T23:04:26.657Z
+assigned: ""
+progress: 0
+tags: []
+---
+
+# Recruit 2–3 pilot users

package/.kanbn/tasks/refine-your-pitch.md

@@ -0,0 +1,9 @@
+---
+created: 2026-04-09T23:04:09.931Z
+updated: 2026-04-09T23:04:09.930Z
+assigned: ""
+progress: 0
+tags: []
+---
+
+# Refine your pitch

package/.kanbn/tasks/write-down-repeated-wording.md

@@ -0,0 +1,9 @@
+---
+created: 2026-04-09T23:03:51.949Z
+updated: 2026-04-09T23:03:51.947Z
+assigned: ""
+progress: 0
+tags: []
+---
+
+# Write down repeated wording

package/.kanbn/tasks/write-the-one-pager.md

@@ -0,0 +1,27 @@
+---
+created: 2026-04-09T22:56:55.687Z
+updated: 2026-04-09T23:14:25.954Z
+assigned: ""
+progress: 1
+tags:
+  - 'Week One'
+due: 2026-04-12T00:00:00.000Z
+completed: 2026-04-10T00:00:00.000Z
+---
+
+# Write the one pager
+
+Before your first mentor call, write one page with only these headings:
+What Norn is
+Who it is for
+What problem it solves
+What people use today instead
+Why those alternatives are painful
+What is already built
+What proof you have so far
+What you need help deciding next
+
+## Comments
+
+- date: 2026-04-09T23:11:55.306Z
+  https://docs.google.com/document/d/10L0HINdu6bvcKK5FhMBthPHUQYFKo7ix8OBFo-CcodU/edit?usp=sharing

package/CHANGELOG.md

@@ -4,6 +4,48 @@ All notable changes to the "Norn" extension will be documented in this file.
 
 ## [Unreleased]
 
+## [1.12.0] - 2026-04-20
+
+### Added
+- **MCP Tool Arguments (Extension + CLI)**:
+  - Added positional MCP tool arguments for `run mcp call <alias> <tool>(...)`, bound in tool-schema order alongside existing named arguments.
+  - Added schema-aware MCP tool resolution before tool execution so required, duplicate, and unknown parameter checks use the server's declared input schema consistently.
+  - Added regression coverage for positional and mixed MCP argument calls in both the fake MCP fixtures and the docker-backed MCP regression.
+
+### Improved
+- **MCP Editor Authoring (VS Code)**:
+  - Added config-scoped `.norn-cache` persistence for listed MCP tool metadata so tool-name completions and parameter hints work from cached discovery data.
+  - Added compact MCP parameter hints with schema-backed descriptions for the active parameter and cleaner tool insertion with `()`.
+  - Suppressed unrelated bare-variable suggestions while typing inside MCP tool argument lists.
+
+### Fixed
+- **MCP Diagnostics and Fixture Metadata**:
+  - Fixed diagnostics so positional MCP tool arguments are not treated like duplicate named arguments.
+  - Enriched the fake MCP regression tool schemas with parameter descriptions so cached hint text stays meaningful during local testing.
+
+## [1.11.0] - 2026-04-19
+
+### Added
+- **Deterministic MCP Tools (Extension + CLI)**:
+  - Added `norn.mcp.json` project config with nearest-file resolution, stdio and HTTP transports, and shared alias-based runtime behavior.
+  - Added `run mcp list <alias>` and `run mcp call <alias> <tool>(name: value)` with variable capture support in both VS Code and the CLI.
+  - Added deterministic MCP session reuse per resolved config file plus alias for the lifetime of a sequence run, including nested and imported sequences.
+  - Added MCP tool output-schema validation, paginated `tools/list` draining, dedicated CLI/report output, response-panel rendering, and regression fixtures for stdio and HTTP MCP servers.
+
+### Improved
+- **Project Setup UX (VS Code)**:
+  - Added `Create SQL starter files` and `Create MCP starter config` commands to the Norn sidebar and Explorer folder context menu.
+  - Added starter `norn.mcp.json` content with inline `_comment` guidance fields for alias usage, local vs remote servers, and optional config fields.
+  - Added MCP-specific response panel styling so MCP steps no longer reuse SQL visuals.
+
+### Fixed
+- **Array Count Path Access**:
+  - Fixed structured path resolution so array values can be asserted with `.count` consistently across captured variables, response-body paths, and MCP tool lists.
+
+### Changed
+- **Website Docs**:
+  - Added MCP documentation to the website, including config setup, `run mcp` usage, CLI parity notes, and VS Code project-setup flows.
+
 ## [1.10.6] - 2026-03-28
 
 ### Fixed