docs(many): add Claude Code configuration and documentation

Add comprehensive AI assistant support:
- CLAUDE.md with repository structure, workflows, and conventions
- .claudeignore to optimize Claude Code performance
- .gitignore entries for AI tool artifacts
- Updated ai-features.md to reference CLAUDE.md
- inst-ai.config.mjs for inst-ai integration

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <[email protected]>

Author: balzss

.claudeignore

@@ -0,0 +1,43 @@
+# Dependencies
+node_modules/
+
+# Build outputs
+coverage/
+__build__/
+.tmp/
+packages/*/lib/
+packages/__docs__/__build__/
+
+# Git
+.git/
+
+# Logs and build info
+*.log
+*.tsbuildinfo
+lerna-debug.log
+npm-debug.log
+
+# Test artifacts
+regression-test/node_modules/
+packages/*/node_modules/
+
+# Images and binary files
+*.png
+*.jpg
+*.jpeg
+*.gif
+*.svg
+*.woff
+*.woff2
+*.ttf
+*.eot
+full_logo.png
+logo.png
+
+# Large changelog
+CHANGELOG.md
+
+# Editor and system files
+.vscode/
+.idea/
+.DS_Store

.gitignore

@@ -17,3 +17,12 @@ __build__
 tsconfig.build.tsbuildinfo
 
 .pnp.*
+
+# AI/LLM tool artifacts (Copilot & Claude)
+.copilot/
+.inst-ai/sessions
+
+# Claude Code specific
+.claude/settings.local.json
+.claude/cache/
+.claude/*.local.*

.inst-ai/templates/jira/README.md

@@ -0,0 +1,45 @@
+# Jira Templates
+
+This directory contains prompt templates for generating Jira tickets from Slack conversations.
+
+## Template Types
+
+### Bug Report Templates
+- `bug-report.extraction.md` - Extracts structured data from bug report conversations
+- `bug-report.generation.md` - Generates final Jira ticket content for bugs
+
+### Feature Request Templates
+- `feature-request.extraction.md` - Extracts structured data from feature request conversations
+- `feature-request.generation.md` - Generates final Jira ticket content for features
+
+## Template Variables
+
+Templates use `{{VARIABLE_NAME}}` syntax for variable substitution:
+
+- `{{CONVERSATION_CONTENT}}` - The full Slack conversation text
+- `{{EXTRACTED_DATA}}` - JSON-formatted extracted data from the conversation
+
+## Customizing Templates
+
+You can override default templates by configuring custom template paths in your `inst-ai.config.js`:
+
+```javascript
+export const config = {
+  jira: {
+    customTemplates: {
+      bugReport: {
+        extraction: './my-templates/custom-bug-extraction.md',
+        generation: './my-templates/custom-bug-generation.md'
+      },
+      featureRequest: {
+        extraction: './my-templates/custom-feature-extraction.md',
+        generation: './my-templates/custom-feature-generation.md'
+      }
+    }
+  }
+}
+```
+
+## Template Format
+
+Templates should be valid Markdown files with embedded prompts for the AI model. The AI will process the template content and generate responses in the expected format (JSON for extraction, ADF JSON for generation).

.inst-ai/templates/jira/bug-report.extraction.md

@@ -0,0 +1,21 @@
+# Bug Report Extraction Template
+
+**Task:** Analyze the conversation and extract entities into a JSON object. Use `null` for missing values.
+
+**Entities:**
+- `component_name`: string | null - The name of the UI component or module mentioned
+- `browser_name`: string | null - Browser where the issue occurs (e.g., "Chrome", "Firefox", "Safari")
+- `os_name`: string | null - Operating system where the issue occurs (e.g., "macOS", "Windows", "Linux")
+- `instui_version`: string | null - InstUI version (e.g., "8.51.0", "v8.51.0"). Look for @instructure/ui-* package versions in package.json, version mentions in conversation, or CodeSandbox dependencies
+- `summary_of_bug`: string - Brief description of the bug
+- `reporter_name`: string - Name of the person reporting the bug
+- `environment_text`: string | null - Additional environment details
+- `steps_to_reproduce`: array | null - List of steps to reproduce the issue
+- `expected_behavior`: string | null - What should happen
+- `actual_behavior`: string | null - What actually happens
+- `workaround`: string | null - Any workarounds mentioned
+
+**Conversation:**
+{{CONVERSATION_CONTENT}}
+
+**JSON Output:**

.inst-ai/templates/jira/bug-report.generation.md

@@ -0,0 +1,27 @@
+# Bug Report Generation Template
+
+**Persona:** Expert Technical Writer creating a developer-ready bug report for Jira.
+
+**Task:** Use the `CONTEXT` to generate a JSON object with a `summary` and an ADF `description`.
+
+**Requirements:**
+- The `summary` must be: `Fix: [<Component Name>] <Brief problem description>`. Use the component_name from extracted data if available, otherwise use a generic name based on the affected area.
+- The `description` must be a valid Atlassian Document Format (ADF) JSON object.
+- Include all relevant technical details from the context.
+- Structure the description with clear sections: Problem, Environment, Steps, Expected/Actual behavior.
+- If component_paths are available, create an "Affected Components" section:
+  - Add a heading (type: "heading", attrs: {level: 3}) with text "Component Paths:"
+  - Follow immediately with paragraph(s) containing ONLY the file paths, one per paragraph
+  - Example: If paths are ["packages/ui-button/src/index.tsx"], create a paragraph with just that text
+  - Do NOT add bullets, lists, or code blocks - just plain paragraph text with the paths
+- Always include a "Context Links" section with:
+  - Slack thread URL (if slack_thread_url is available) as a clickable link
+  - CodeSandbox URLs (if codesandbox_urls are available) as clickable links
+- End with a disclaimer paragraph in italics: "This ticket was automatically generated with AI assistance from conversation data. Please review and update as needed."
+
+**CONTEXT:**
+{{EXTRACTED_DATA}}
+
+**Your Turn (Use the CONTEXT provided above):**
+**IMPORTANT:** Return ONLY valid JSON. Every property must be followed by a comma except the last one in an object or array. Double-check all commas before responding.
+**Output JSON:**

.inst-ai/templates/jira/feature-request.extraction.md

@@ -0,0 +1,18 @@
+# Feature Request Extraction Template
+
+**Task:** Analyze the conversation and extract entities into a JSON object. Use `null` for missing values.
+
+**Entities:**
+- `feature_name`: string | null - The name of the requested feature
+- `requestor_name`: string - Name of the person requesting the feature
+- `business_justification`: string | null - Why this feature is needed
+- `proposed_solution`: string | null - How the feature should work
+- `affected_components`: array | null - Components that would be affected
+- `priority_level`: string | null - Urgency or priority mentioned
+- `success_criteria`: string | null - How to measure success
+- `alternative_solutions`: array | null - Other approaches considered
+
+**Conversation:**
+{{CONVERSATION_CONTENT}}
+
+**JSON Output:**

.inst-ai/templates/jira/feature-request.generation.md

@@ -0,0 +1,27 @@
+# Feature Request Generation Template
+
+**Persona:** Product Manager creating a comprehensive feature request for Jira.
+
+**Task:** Use the `CONTEXT` to generate a JSON object with a `summary` and an ADF `description`.
+
+**Requirements:**
+- The `summary` must be: `Feature: [<Component/Area Name>] <Brief description>`. Use the feature_name or affected_components from extracted data if available, otherwise use a generic name based on the affected area.
+- The `description` must be a valid Atlassian Document Format (ADF) JSON object.
+- Include business justification, proposed solution, and success criteria.
+- Structure the description with clear sections: Overview, Business Case, Requirements, Acceptance Criteria.
+- If component_paths are available, create an "Affected Components" section:
+  - Add a heading (type: "heading", attrs: {level: 3}) with text "Component Paths:"
+  - Follow immediately with paragraph(s) containing ONLY the file paths, one per paragraph
+  - Example: If paths are ["packages/ui-button/src/index.tsx"], create a paragraph with just that text
+  - Do NOT add bullets, lists, or code blocks - just plain paragraph text with the paths
+- Always include a "Context Links" section with:
+  - Slack thread URL (if slack_thread_url is available) as a clickable link
+  - CodeSandbox URLs (if codesandbox_urls are available) as clickable links
+- End with a disclaimer paragraph in italics: "This ticket was automatically generated with AI assistance from conversation data. Please review and update as needed."
+
+**CONTEXT:**
+{{EXTRACTED_DATA}}
+
+**Your Turn (Use the CONTEXT provided above):**
+**IMPORTANT:** Return ONLY valid JSON. Every property must be followed by a comma except the last one in an object or array. Double-check all commas before responding.
+**Output JSON:**