From 52d12a28b7e148af381cda8e80f10406c7e73189 Mon Sep 17 00:00:00 2001
From: Florian BRUNIAUX <florian@bruniaux.com>
Date: Thu, 5 Mar 2026 16:18:24 +0100
Subject: [PATCH] =?UTF-8?q?release:=20v3.30.2=20=E2=80=94=20issue-triage?=
 =?UTF-8?q?=20skill,=20design-reference-file,=20Conductor=20docs?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

New templates:
- examples/skills/issue-triage/ — 3-phase issue backlog management with
  Jaccard duplicate detection, risk classification, and validated actions
- examples/claude-md/design-reference-file.md — brand-book + ui-kit
  pattern for consistent UI generation across sessions

Resource evaluation:
- docs/resource-evaluations/075-paillard-design-system-first-website.md
  (Boris Paillard, mixt.care, score 3/5)

Docs update:
- guide/third-party-tools.md — Conductor section enriched with verified
  features (Next Workspace, Manual Mode, GitHub CI integration, BMAD pattern)

Version bump: 3.30.1 → 3.30.2 (synced across README, cheatsheet, guide, reference.yaml)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---
 CHANGELOG.md                                  |   2 +
 README.md                                     |   4 +-
 VERSION                                       |   2 +-
 ...75-paillard-design-system-first-website.md |  76 +++
 examples/README.md                            |   5 +-
 examples/claude-md/design-reference-file.md   | 176 +++++++
 examples/skills/issue-triage/SKILL.md         | 450 ++++++++++++++++++
 .../issue-triage/templates/issue-comment.md   | 107 +++++
 guide/cheatsheet.md                           |   4 +-
 guide/ultimate-guide.md                       |   6 +-
 machine-readable/reference.yaml               |  22 +-
 11 files changed, 837 insertions(+), 17 deletions(-)
 create mode 100644 docs/resource-evaluations/075-paillard-design-system-first-website.md
 create mode 100644 examples/claude-md/design-reference-file.md
 create mode 100644 examples/skills/issue-triage/SKILL.md
 create mode 100644 examples/skills/issue-triage/templates/issue-comment.md

diff --git a/CHANGELOG.md b/CHANGELOG.md
index e3c3bc3..8edf11a 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -6,6 +6,8 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 
 ## [Unreleased]
 
+## [3.30.2] - 2026-03-05
+
 ### Documentation
 
 - **Conductor section enriched** (`guide/third-party-tools.md:376`) — section rewritten from 5 generic bullets to 6 structured subsections, verified against official changelog (conductor.build). Added: workspace status system (backlog → done, v0.35.0), Next Workspace button for agent queue navigation (v0.36.4), turn-by-turn diff viewer (v0.22.0), Manual Mode built-in editor replacing VSCode for quick edits (v0.37.0), full GitHub/CI integration details (Actions tab v0.33.2, failing CI → Claude auto-fix v0.12.0, PR workflow with `⌘⇧P`), Linear deeplinks, Codex support alongside Claude, Melty Labs attribution. Community workflow pattern added: 5+ parallel features across multiple repos, BMAD + Conductor combo for spec-driven development (user-reported, unverified claim clearly marked).
diff --git a/README.md b/README.md
index 0a41866..6a560fb 100644
--- a/README.md
+++ b/README.md
@@ -6,7 +6,7 @@
 
 <p align="center">
   <a href="https://github.com/FlorianBruniaux/claude-code-ultimate-guide/stargazers"><img src="https://img.shields.io/github/stars/FlorianBruniaux/claude-code-ultimate-guide?style=for-the-badge" alt="Stars"/></a>
-  <a href="./CHANGELOG.md"><img src="https://img.shields.io/badge/Updated-Mar_4,_2026_·_v3.30.1-brightgreen?style=for-the-badge" alt="Last Update"/></a>
+  <a href="./CHANGELOG.md"><img src="https://img.shields.io/badge/Updated-Mar_5,_2026_·_v3.30.2-brightgreen?style=for-the-badge" alt="Last Update"/></a>
   <a href="./quiz/"><img src="https://img.shields.io/badge/Quiz-274_questions-orange?style=for-the-badge" alt="Quiz"/></a>
   <a href="./examples/"><img src="https://img.shields.io/badge/Templates-176-green?style=for-the-badge" alt="Templates"/></a>
   <a href="./guide/security-hardening.md"><img src="https://img.shields.io/badge/🛡️_Threat_DB-24_CVEs_·_655_malicious_skills-red?style=for-the-badge" alt="Threat Database"/></a>
@@ -846,7 +846,7 @@ See [CONTRIBUTING.md](./CONTRIBUTING.md) for guidelines.
 
 ---
 
-*Version 3.30.1 | Updated daily · Mar 4, 2026 | Crafted with Claude*
+*Version 3.30.2 | Updated daily · Mar 5, 2026 | Crafted with Claude*
 
 <!-- SEO Keywords -->
 <!-- claude code, claude code tutorial, anthropic cli, ai coding assistant, claude code mcp,
diff --git a/VERSION b/VERSION
index 72bde0a..aaa0fde 100644
--- a/VERSION
+++ b/VERSION
@@ -1 +1 @@
-3.30.1
+3.30.2
diff --git a/docs/resource-evaluations/075-paillard-design-system-first-website.md b/docs/resource-evaluations/075-paillard-design-system-first-website.md
new file mode 100644
index 0000000..08d7e9f
--- /dev/null
+++ b/docs/resource-evaluations/075-paillard-design-system-first-website.md
@@ -0,0 +1,76 @@
+# Resource Evaluation: Boris Paillard — "Son site custom en 2h avec Claude Code"
+
+**Date**: 2026-03-05
+**Source**: [LinkedIn Pulse](https://www.linkedin.com/pulse/son-site-custom-en-2h-avec-claude-code-m%C3%A9thodes-et-prompts-paillard-7q8je/)
+**Type**: Tutorial / Workflow walkthrough
+**Author**: Boris Paillard — ex-Le Wagon instructor, co-founder mixt.care
+**Score**: 3/5
+
+---
+
+## Summary
+
+Workflow for building a custom website in ~2h using Claude Code, centered on a "design-first" approach. Core thesis: Claude Code executes your design taste — it doesn't replace it. You must define a design system before prompting, or every generated page drifts.
+
+Four-step method with concrete prompts published in the article:
+1. **brand-book.html** — color palette with semantic roles, fonts, CSS variables
+2. **ui-kit.html** — base components documented with Tailwind CSS
+3. **Full site development** — scroll animations, sticky images, statement footers
+4. **Conditional forms** (bonus) — JSON-driven multi-step forms with Google Apps Script
+
+Concrete project: mixt.care (personalized dermatology). Stack: Tailwind CSS + vanilla JS (Intersection Observer).
+
+---
+
+## Score Justification
+
+**3/5 — Relevant, integrate as field example.**
+
+The four-step methodology is structurally sound and the 4 prompts are specific and usable. The core insight (HTML reference files as persistent Claude Code context) is not documented in the guide and is genuinely reusable.
+
+However: no WCAG/accessibility coverage, no differentiation from Cursor/Copilot, no GitHub repo to verify results, and no discussion of maintainability past the MVP. The guide's technical sections need more rigorous sources.
+
+---
+
+## Gap Identified → Action Taken
+
+**Gap**: The guide had no documentation of the "Design Reference File" pattern — keeping `brand-book.html` and `ui-kit.html` at the project root as permanent context files for Claude Code. This pattern ensures design coherence across all generated pages without re-prompting.
+
+**Action**: Added `examples/claude-md/design-reference-file.md` — a standalone template for the pattern, with:
+- Recommended project structure
+- CLAUDE.md snippet to activate automatic design system reference
+- Prompt for brand-book.html with WCAG audit embedded
+- Prompt for ui-kit.html documentation
+- Color audit prompt (WCAG 2.1 compliance, color blindness simulation, fix suggestions)
+
+No changes to `ultimate-guide.md` — the pattern is documented as an example, not promoted to a guide section. Score does not justify central placement.
+
+---
+
+## What the Article Does NOT Cover
+
+- WCAG accessibility / contrast ratios for the generated palette
+- Maintainability after the 2h MVP (technical debt, component evolution)
+- Why Claude Code specifically vs. Cursor or Copilot
+- No GitHub repo or live code to verify the claimed output
+
+---
+
+## Fact-Check
+
+| Claim | Status |
+|-------|--------|
+| Author: Boris Paillard | Confirmed — LinkedIn byline |
+| Date: March 5, 2026 | Confirmed — `datePublished: 2026-03-05T08:14:35Z` |
+| Project: mixt.care (dermatology) | Confirmed — mentioned explicitly |
+| 2h timeframe | Confirmed — title + body |
+| Font sizes: 6rem desktop / 3rem mobile | Confirmed — in footer prompt |
+| Animations: 0.6s + 100ms stagger | Confirmed — in Intersection Observer prompt |
+| 4 prompts published in article | Confirmed — verified on second fetch |
+| "Co-founder Le Wagon" | Correction — article says "launched a workshop 10 years ago", not co-founder |
+
+---
+
+## Decision
+
+**Integrate as field example.** The "Design Reference File" pattern (brand-book.html + ui-kit.html as permanent project context) is the one novel, reusable insight. Extract as a standalone example template. Do not cite as a primary technical source.
diff --git a/examples/README.md b/examples/README.md
index 21d627e..08c59fd 100644
--- a/examples/README.md
+++ b/examples/README.md
@@ -17,7 +17,7 @@ Annotated templates that teach you **why** patterns work, not just how to config
 | [`agents/`](./agents/) | Custom AI personas for specialized tasks | 9 |
 | [`commands/`](./commands/) | Slash commands (workflow automation) | 26 |
 | [`hooks/`](./hooks/) | Event-driven security & automation scripts | 31 |
-| [`skills/`](./skills/) | Reusable knowledge modules — [9 on SkillHub](https://skills.palebluedot.live/owner/FlorianBruniaux) | 14 |
+| [`skills/`](./skills/) | Reusable knowledge modules — [9 on SkillHub](https://skills.palebluedot.live/owner/FlorianBruniaux) | 15 |
 | [`claude-md/`](./claude-md/) | CLAUDE.md configuration profiles | 6 |
 | [`config/`](./config/) | Settings, MCP, git templates | 5 |
 | [`memory/`](./memory/) | CLAUDE.md memory file templates | 2 |
@@ -67,7 +67,7 @@ Annotated templates that teach you **why** patterns work, not just how to config
 | [implementer.md](./agents/implementer.md) | Mechanical execution — bounded scope | Haiku |
 | [architecture-reviewer.md](./agents/architecture-reviewer.md) | Architecture & design review — read-only | Opus |
 
-### Skills (14) — [9 on SkillHub](https://skills.palebluedot.live/owner/FlorianBruniaux)
+### Skills (15) — [9 on SkillHub](https://skills.palebluedot.live/owner/FlorianBruniaux)
 
 | File | Purpose |
 |------|---------|
@@ -85,6 +85,7 @@ Annotated templates that teach you **why** patterns work, not just how to config
 | [guide-recap/](./skills/guide-recap/) | Transform CHANGELOG entries into social content (LinkedIn, Twitter/X, Slack) |
 | [release-notes-generator/](./skills/release-notes-generator/) | Generate release notes in 3 formats from git commits |
 | [pr-triage/](./skills/pr-triage/) | 3-phase PR backlog management (audit, deep review, validated comments) |
+| [issue-triage/](./skills/issue-triage/) | 3-phase issue backlog management (audit, deep analysis, validated actions) |
 
 ### Commands (26)
 
diff --git a/examples/claude-md/design-reference-file.md b/examples/claude-md/design-reference-file.md
new file mode 100644
index 0000000..4b0f340
--- /dev/null
+++ b/examples/claude-md/design-reference-file.md
@@ -0,0 +1,176 @@
+---
+title: "Design Reference File"
+description: "Keep brand-book.html and ui-kit.html at project root as permanent Claude Code context for consistent UI generation"
+tags: [design-system, frontend, web, ui, consistency, brand, color-palette, tailwind]
+---
+
+# Design Reference File — CLAUDE.md Pattern
+
+Keep `brand-book.html` and `ui-kit.html` at the project root as permanent context files. Claude Code reads them before generating any UI — every new page inherits your design system automatically.
+
+Inspired by Boris Paillard's workflow (mixt.care, March 2026): once the design system is in place, new pages take 5 minutes instead of 30.
+
+## The Problem
+
+When building a website with Claude Code across multiple sessions, every new page risks drifting from the design — wrong colors, inconsistent typography, new component variants. Re-stating design constraints in each prompt is repetitive and unreliable.
+
+## The Solution
+
+Two HTML files at the project root act as a design memory Claude can read at any time:
+
+- `brand-book.html` — color palette with semantic roles, fonts, CSS variables, WCAG contrast ratios
+- `ui-kit.html` — documented components (buttons, forms, trust bars, section labels)
+
+One CLAUDE.md instruction makes Claude reference them before every UI task.
+
+## Project Structure
+
+```
+project/
+├── brand-book.html     # Color palette + typography + CSS variables (permanent reference)
+├── ui-kit.html         # Component library documented with Tailwind
+├── CLAUDE.md           # Design system instruction below
+├── src/
+│   ├── styles/
+│   │   └── tokens.css  # CSS variables extracted from brand-book.html
+│   └── components/
+└── ...
+```
+
+## CLAUDE.md Snippet
+
+Add this to your project-level `CLAUDE.md`:
+
+```markdown
+## Design System
+
+Always read `brand-book.html` and `ui-kit.html` before generating any UI component, page, or layout.
+
+Rules:
+- Never hardcode colors or font sizes — use CSS variables from brand-book.html
+- Reuse components from ui-kit.html before creating new variants
+- New pages must use the same design tokens (--color-primary, --font-primary, etc.)
+- If a requested design element is not in the UI kit, document it in ui-kit.html after creating it
+```
+
+## Step 1 — Generate brand-book.html
+
+```
+Create a brand-book.html file at the project root.
+
+For each color in my palette, display a card showing:
+- Color swatch (120×80px block)
+- Name, hex code, RGB values, HSL values
+- CSS variable name (e.g. --color-primary)
+- Semantic role: PRIMARY | DARK | LIGHT | ACCENT | NEUTRAL
+- Usage rule in plain text (e.g. "CTAs, buttons, active links")
+- WCAG contrast ratio vs white: X.X:1 — AA PASS/FAIL — AAA PASS/FAIL
+- WCAG contrast ratio vs black: X.X:1 — AA PASS/FAIL — AAA PASS/FAIL
+
+My palette:
+[LIST YOUR COLORS AND ROLES HERE]
+
+My fonts:
+[LIST YOUR FONTS HERE]
+
+Include a type scale section: 12px / 16px / 20px / 24px / 32px / 48px / 64px / 96px with rem equivalents.
+
+At the bottom, output a copyable <style> block with all CSS variables (:root { ... }).
+
+Style brand-book.html itself using these variables — it should demonstrate the design system.
+```
+
+## Step 2 — Generate ui-kit.html
+
+```
+Build ui-kit.html documenting my base components using Tailwind and the CSS variables from brand-book.html.
+
+Include:
+- Buttons: primary/secondary/ghost variants, 3 sizes (sm/md/lg)
+- Trust bars with vertical color separators
+- Section labels with horizontal rules (uppercase, letter-spacing)
+- Typography specimens: h1–h4, body, caption — using --font-primary and --font-secondary
+- Color swatches grid with contrast ratios displayed
+- Form elements: input, select, textarea in default/focus/error states
+
+Each component should show: component name, usage notes, and the Tailwind classes used.
+Reference CSS variables from brand-book.html — no hardcoded values.
+```
+
+## Step 3 — WCAG Color Audit (optional but recommended)
+
+Run this after generating the brand book to catch accessibility issues before building:
+
+```
+Audit the color palette in brand-book.html for WCAG 2.1 compliance:
+
+1. For all likely foreground/background combinations, calculate exact contrast ratios (2 decimal places)
+2. Flag pairs that fail WCAG AA: 4.5:1 for normal text, 3:1 for large text and UI components
+3. Simulate deuteranopia, protanopia, tritanopia — flag problematic combinations
+4. For each failing pair, suggest the minimum hex adjustment to pass AA while staying within 10% hue deviation
+5. Output: markdown table with columns: Combination | Ratio | AA Normal | AA Large | AAA | Color Blind Safe
+
+External references:
+- WebAIM Contrast Checker: https://webaim.org/resources/contrastchecker/
+- ColorOracle (color blindness simulator): https://colororacle.org/
+```
+
+## Step 4 — Scroll Animations (optional)
+
+```
+Add scroll animations to all cards and section titles using vanilla JS Intersection Observer.
+Elements should fade in + slide up with 0.6s ease-out transitions, 100ms stagger between items.
+Apply only to elements with [data-animate] attribute — opt-in, not global.
+Do not use any animation library — vanilla JS only.
+```
+
+## Example — mixt.care Color Palette
+
+Well-structured palette with semantic roles and CSS variables:
+
+```css
+:root {
+  /* Colors */
+  --color-primary: #5C1A2E;   /* Deep bordeaux — CTAs, buttons, active links */
+  --color-dark: #3D2B1F;      /* Warm brown — body text, dark backgrounds */
+  --color-accent: #B87333;    /* Copper — hover states, secondary highlights */
+  --color-secondary: #6B5B8E; /* Soft violet — badges, illustrations */
+  --color-light: #F5F0EA;     /* Cream — section backgrounds, cards */
+
+  /* Typography */
+  --font-primary: 'Geist', sans-serif;       /* Body text */
+  --font-secondary: 'Newsreader', serif;     /* Emphasis, quotes */
+  --font-display: 'Fraunces', serif;         /* Logo, display headings only */
+
+  /* Type scale */
+  --text-xs: 0.75rem;
+  --text-sm: 0.875rem;
+  --text-base: 1rem;
+  --text-lg: 1.25rem;
+  --text-xl: 1.5rem;
+  --text-2xl: 2rem;
+  --text-3xl: 3rem;
+  --text-display: 6rem;        /* Desktop statement footer */
+  --text-display-mobile: 3rem;
+}
+```
+
+**WCAG notes for this palette** (estimated):
+- Bordeaux on cream: ~8.2:1 — AA PASS
+- Brown on cream: ~9.1:1 — AA PASS
+- White on bordeaux: ~8.2:1 — AA PASS
+- Copper on white: ~3.1:1 — AA FAIL for normal text (use only for UI components or darken to #9B5F20)
+- Violet on cream: ~4.8:1 — AA PASS
+
+## When to Use This Pattern
+
+- Building a marketing site, landing page, or product website with Claude Code
+- When you need consistent design across multiple pages generated in different sessions
+- Before starting UI work on any project with established brand guidelines
+- Replacing manual design system documentation with an interactive, self-demonstrating reference
+
+## Limitations
+
+- Effective for static/marketing sites and MVPs — complex component libraries need a proper design system tool (Storybook, Figma)
+- The 2h timeframe assumes you know your design intent before prompting (colors, fonts, layout direction)
+- WCAG audit from Claude is estimated — verify critical pairs with WebAIM or a dedicated tool
\ No newline at end of file
diff --git a/examples/skills/issue-triage/SKILL.md b/examples/skills/issue-triage/SKILL.md
new file mode 100644
index 0000000..ec9a38f
--- /dev/null
+++ b/examples/skills/issue-triage/SKILL.md
@@ -0,0 +1,450 @@
+---
+name: issue-triage
+description: >
+  3-phase issue backlog management: audit open issues, deep analyze selected ones, draft and execute
+  triage actions with mandatory validation. Args: "all" to analyze all, issue numbers to focus
+  (e.g. "42 57"), "en"/"fr" for language, no arg = audit only in French.
+tags: [github, issue, triage, maintainer, multi-agent]
+---
+
+# Issue Triage
+
+3-phase workflow for maintainers: automated audit of all open issues, opt-in deep analysis via parallel agents, and validated triage actions (comments, labels, closures).
+
+## When to Use This Skill
+
+| Skill | Usage | Output |
+|-------|-------|--------|
+| `/issue-triage` | Sort, analyze, and act on an issue backlog | Triage tables + analysis + executed actions |
+| `/pr-triage` | Sort, review, and comment on a PR backlog | Triage table + reviews + posted comments |
+
+**Triggers**:
+- Manually: `/issue-triage` or `/issue-triage all` or `/issue-triage 42 57`
+- Proactively: when >10 open issues without label, or stale issues >30 days detected
+
+---
+
+## Language
+
+- Check the argument passed to the skill
+- If `en` or `english` → tables and summary in English
+- If `fr`, `french`, or no argument → French (default)
+- Note: GitHub comments and labels (Phase 3) are ALWAYS in English (international audience)
+
+---
+
+## Configuration
+
+Thresholds used throughout the workflow. Edit to match your project:
+
+| Parameter | Default | Description |
+|-----------|---------|-------------|
+| `staleness_days` | 30 | Days without activity before flagging as stale |
+| `very_stale_days` | 90 | Days without activity before flagging as very stale |
+| `jaccard_threshold` | 60% | Minimum Jaccard similarity to flag two issues as duplicates |
+| `closed_compare_count` | 20 | Number of recent closed issues to compare for duplicate detection |
+| `open_limit` | 100 | Maximum open issues to fetch and analyze |
+
+---
+
+## Preconditions
+
+```bash
+git rev-parse --is-inside-work-tree
+gh auth status
+```
+
+If either fails, stop and explain what is missing.
+
+---
+
+## Phase 1 — Audit (always executed)
+
+### Data Gathering (parallel commands)
+
+```bash
+# Repo identity
+gh repo view --json nameWithOwner -q .nameWithOwner
+
+# Open issues (exclude PRs, limit 100)
+gh issue list --state open --limit 100 \
+  --json number,title,author,createdAt,updatedAt,labels,body,comments,assignees,milestone
+
+# Recent closed issues (for duplicate detection)
+gh issue list --state closed --limit 20 \
+  --json number,title,body,labels,stateReason
+
+# Open PRs (bodies for cross-reference detection)
+gh pr list --state open --limit 50 --json number,title,body
+
+# Collaborators (to distinguish reporter types)
+gh api "repos/{owner}/{repo}/collaborators" --jq '.[].login'
+```
+
+**Collaborators fallback**: if `gh api .../collaborators` returns 403/404:
+```bash
+# Extract authors from last 10 merged PRs
+gh pr list --state merged --limit 10 --json author --jq '.[].author.login' | sort -u
+```
+If still ambiguous, ask via `AskUserQuestion`.
+
+**Note**: `comments` field in `gh issue list --json comments` returns the count, not content. For Phase 2, fetch full content separately: `gh issue view {num} --json comments`.
+
+### Analysis Dimensions
+
+Run all 6 dimensions for each open issue:
+
+#### 1. Categorization
+
+Classify each issue by reading `title` + first 200 chars of `body`:
+
+| Category | Label | Criteria |
+|----------|-------|----------|
+| Bug | `bug` | Describes broken behavior, unexpected output, crash |
+| Feature Request | `enhancement` | Asks for new functionality |
+| Question / Support | `question` | User asking how something works |
+| Documentation | `documentation` | Missing or incorrect docs |
+| Out of Scope | `wontfix` | Clearly outside project boundaries |
+| Unclear | `needs-info` | Body empty, too vague to categorize |
+
+If body is empty → category is always `Unclear` (never assume).
+
+#### 2. Cross-reference to PRs
+
+Scan each open PR body for references to the issue number:
+- Patterns: `fixes #N`, `closes #N`, `resolves #N`, `fix #N`, `close #N` (case-insensitive, `N` = issue number)
+- Use regex locally on the `body` fields already fetched — do NOT make N additional API calls
+- If found: flag issue as "PR-linked" with PR number
+
+#### 3. Duplicate Detection via Jaccard Similarity
+
+**Algorithm (self-contained — no external library)**:
+
+For each open issue, compute Jaccard similarity against all other open issues AND the 20 most recent closed issues.
+
+```
+Step 1 — Normalize title + first 300 chars of body:
+  - Lowercase the full text
+  - Strip category prefixes: "feat:", "fix:", "bug:", "chore:", "docs:", "test:", "refactor:"
+  - Remove punctuation: .,!?;:'"()[]{}-_/\@#
+
+Step 2 — Tokenize:
+  - Split on whitespace
+  - Remove stop words: the a an is in on to for of and or with this that it can not no be
+  - Remove tokens shorter than 3 characters
+
+Step 3 — Compute Jaccard:
+  tokens_A = set of tokens from issue A
+  tokens_B = set of tokens from issue B
+  jaccard = |tokens_A ∩ tokens_B| / |tokens_A ∪ tokens_B|
+
+Step 4 — Flag:
+  - If jaccard >= 0.60: mark as potential duplicate
+  - Report: "Similar to #N (Jaccard: 0.72)"
+  - Keep the OLDER issue as canonical; newer = duplicate candidate
+```
+
+Jaccard is computed at runtime using the fetched data — no API calls beyond Phase 1 gather.
+
+#### 4. Risk Classification
+
+Assign Red / Yellow / Green based on signals in title + body:
+
+| Level | Color | Criteria |
+|-------|-------|----------|
+| Critical | Red | Security vulnerability, data loss, regression blocking users, crash in production |
+| Needs Attention | Yellow | Missing validation, performance degradation, breaking change undocumented, Unclear with no response for >7 days |
+| Normal | Green | Everything else |
+
+#### 5. Staleness
+
+| Status | Criterion |
+|--------|-----------|
+| Active | Updated within 30 days |
+| Stale | No activity 30–90 days |
+| Very Stale | No activity >90 days |
+
+Use `updatedAt` field. Staleness does NOT depend on comments count — a commented-on issue with old `updatedAt` is still stale.
+
+#### 6. Recommendations
+
+One recommended action per issue:
+
+| Situation | Action |
+|-----------|--------|
+| Category = Unclear, body empty | Comment requesting details |
+| Jaccard >= 0.60 with known issue | Close as duplicate, link original |
+| Very stale + no assignee | Comment requesting status, suggest close |
+| Risk = Red | Pin to top of triage, escalate immediately |
+| Category = OOS | Close with explanation |
+| PR-linked | No action needed (tracked via PR) |
+| Normal + labeled | No action needed |
+
+### Output — Triage Tables
+
+```
+## Open Issues ({count})
+
+### Critical — Immediate Attention (Risk: Red)
+| # | Title | Category | Reporter | Days Open | Action |
+|---|-------|----------|----------|-----------|--------|
+
+### PR-Linked (tracked in open PRs)
+| # | Title | Category | PR | Days Open |
+|---|-------|----------|----|-----------|
+
+### Active Issues
+| # | Title | Category | Labels | Reporter | Days | Action |
+|---|-------|----------|--------|----------|------|--------|
+
+### Duplicate Candidates
+| # | Title | Similar To | Jaccard | Action |
+|---|-------|------------|---------|--------|
+
+### Stale Issues
+| # | Title | Category | Last Activity | Reporter | Action |
+|---|-------|----------|---------------|----------|--------|
+
+### Summary
+- Total open: {N}
+- Critical (Red): {count}
+- PR-linked: {count}
+- Duplicate candidates: {count}
+- Stale (30–90d): {count}
+- Very stale (>90d): {count}
+- Unlabeled: {count}
+- Recommended actions: {comment: N, label: N, close: N}
+```
+
+0 issues → display `No open issues.` and stop.
+
+**Protection rules** (apply to all phases):
+- Never close an issue authored by a collaborator without explicit user confirmation
+- Never re-label an issue that already has labels (only add missing labels)
+- If body is empty → always request details before any other action
+- Never auto-close a Red issue without user confirmation
+
+### Automatic Copy
+
+After displaying the triage tables, copy to clipboard using platform-appropriate command:
+
+```bash
+UNAME=$(uname -s)
+if [ "$UNAME" = "Darwin" ]; then
+  pbcopy <<'EOF'
+{full triage tables}
+EOF
+elif command -v xclip &>/dev/null; then
+  echo "{full triage tables}" | xclip -selection clipboard
+elif command -v wl-copy &>/dev/null; then
+  echo "{full triage tables}" | wl-copy
+elif command -v clip.exe &>/dev/null; then
+  echo "{full triage tables}" | clip.exe
+fi
+```
+
+Confirm: `Triage tables copied to clipboard.` (EN) / `Tableaux copiés dans le presse-papier.` (FR)
+
+---
+
+## Phase 2 — Deep Analysis (opt-in)
+
+### Issue Selection
+
+**If argument passed**:
+- `"all"` → all issues with recommended actions
+- Numbers (`"42 57"`) → only those issues
+- No argument → propose via `AskUserQuestion`
+
+**If no argument**, display:
+
+```
+question: "Which issues do you want to analyze in depth?"
+header: "Deep Analysis"
+multiSelect: true
+options:
+  - label: "All ({N} issues with recommended actions)"
+    description: "Launch parallel analysis agents for each actionable issue"
+  - label: "Critical only ({M} Red issues)"
+    description: "Focus on high-risk issues requiring immediate action"
+  - label: "Duplicate candidates ({K} issues)"
+    description: "Verify Jaccard similarity with full body + comments"
+  - label: "Stale only ({J} stale issues)"
+    description: "Decide which stale issues to close vs. revive"
+  - label: "Skip"
+    description: "Stop here — audit only"
+```
+
+If "Skip" → end workflow.
+
+### Executing Analysis
+
+For each selected issue, launch an analysis agent via **Task tool in parallel**:
+
+```
+subagent_type: general
+model: sonnet
+prompt: |
+  Analyze GitHub issue #{num}: "{title}"
+
+  **Metadata**: Category={category}, Risk={risk}, Days open={days}, Labels={labels}
+  **Reporter**: @{author} ({collaborator? "collaborator" : "external"})
+  **Assignees**: {assignees or "none"}
+
+  **Body**:
+  {body}
+
+  **Comments** (fetch via: gh issue view {num} --json comments):
+  {comments[].body — truncate at 5000 chars total}
+
+  **Duplicate candidates**: {jaccard_results or "none found"}
+  **Linked PRs**: {pr_refs or "none"}
+
+  Tasks:
+  1. Verify the category assigned in Phase 1 (correct? suggest alternative if not)
+  2. If duplicate candidate: confirm or deny similarity with rationale
+  3. If Unclear/needs-info: identify exactly what information is missing
+  4. Suggest the most appropriate action with exact text if a comment is needed
+  5. Estimate effort to fix if it's a Bug or Feature Request (XS/S/M/L/XL)
+
+  Return structured output:
+  ### Verification
+  ### Duplicate Analysis
+  ### Missing Information
+  ### Recommended Action
+  ### Effort Estimate
+```
+
+**Fallback if parallel agents unavailable**: run analysis sequentially, one issue at a time. Notify user: `Running sequential analysis (parallel agents not available).`
+
+Fetch full comments via:
+```bash
+gh issue view {num} --json comments --jq '.comments[].body'
+```
+
+Aggregate all reports. Display a summary after all analyses complete.
+
+---
+
+## Phase 3 — Actions (mandatory validation)
+
+### Draft Generation
+
+For each analyzed issue, generate the appropriate action using the template `templates/issue-comment.md`.
+
+**3 action types**:
+
+| Type | Command | When |
+|------|---------|------|
+| Comment | `gh issue comment {num} --body-file -` | Needs info, stale ping, OOS explanation |
+| Label | `gh issue edit {num} --add-label "{label}"` | Unlabeled issue with clear category |
+| Close | `gh issue close {num} --reason "not planned"` | Duplicate, OOS, very stale |
+
+**Rules**:
+- Language for comments: **English** (international audience)
+- Labels added: use existing repo labels only (fetch with `gh label list`)
+- Close reason: `"not planned"` for OOS/duplicate, `"completed"` only if a fix was merged
+- Never post a comment AND close in the same action without user seeing both drafts
+- Always attach a comment when closing (explain why)
+
+### Display and Validation
+
+**Display ALL drafted actions** in format:
+
+```
+---
+### Draft — Issue #{num}: {title}
+
+**Action**: {Comment / Label / Close + Comment}
+**Reason**: {1 sentence}
+
+{full comment text if applicable}
+
+---
+```
+
+Then request validation via `AskUserQuestion`:
+
+```
+question: "These actions are ready. Which ones do you want to execute?"
+header: "Execute Triage Actions"
+multiSelect: true
+options:
+  - label: "All ({N} actions)"
+    description: "Execute all drafted triage actions"
+  - label: "Issue #{x} — {title_truncated} ({action_type})"
+    description: "Execute only this action"
+  - label: "None"
+    description: "Cancel — execute nothing"
+```
+
+(Generate one option per issue + "All" + "None")
+
+### Execution
+
+For each validated action:
+
+```bash
+# Comment
+gh issue comment {num} --body-file - <<'TRIAGE_EOF'
+{comment}
+TRIAGE_EOF
+
+# Label
+gh issue edit {num} --add-label "{label}"
+
+# Close with comment
+gh issue comment {num} --body-file - <<'TRIAGE_EOF'
+{close comment}
+TRIAGE_EOF
+gh issue close {num} --reason "not planned"
+```
+
+Confirm each action: `Action executed on issue #{num}: {title}`
+
+If "None" → `No actions executed. Workflow complete.`
+
+---
+
+## Edge Cases
+
+| Situation | Behavior |
+|-----------|----------|
+| 0 open issues | Display `No open issues.` + stop |
+| Body empty | Category = Unclear, action = request details, never assume |
+| Collaborator as reporter | Protect from auto-close, flag explicitly in table |
+| Jaccard inconclusive (0.55–0.65) | Flag as "possible duplicate — verify manually" |
+| Label not in repo | Skip label action, notify user to create the label first |
+| Issue already closed during workflow | Skip silently, note in summary |
+| `gh api .../collaborators` 403/404 | Fallback to last 10 merged PR authors |
+| Parallel agents unavailable | Run sequential analysis, notify user |
+| Very large body (>5000 chars) | Truncate to 5000 chars with `[truncated]` note |
+| Milestone assigned | Include in table, never close milestoned issues without confirmation |
+
+---
+
+## Notes
+
+- Always derive owner/repo via `gh repo view`, never hardcode
+- Use `gh` CLI (not `curl` GitHub API) except for collaborators list
+- `comments` in `gh issue list --json comments` = count only; full content requires `gh issue view {num} --json comments`
+- Never execute any action without explicit user validation in chat
+- Drafted actions must be visible BEFORE any `gh issue comment` or `gh issue close`
+- Jaccard is computed locally — no external API, no library, pure set operations on fetched data
+- Signature on all comments: `*Triaged via Claude Code /issue-triage*`
+
+---
+
+## Related: /pr-triage
+
+| | `/issue-triage` | `/pr-triage` |
+|--|----------------|--------------|
+| **Scope** | Issue backlog | PR backlog |
+| **Use when** | Catching up on reporter feedback, periodic issue cleanup | Catching up after PR accumulation |
+| **Phases** | 3 (audit + deep analysis + actions) | 3 (audit + deep review + comments) |
+| **Agents** | Parallel sub-agents per issue | Parallel sub-agents per PR |
+| **Duplicate detection** | Jaccard similarity on title+body | File overlap % between PRs |
+| **Actions** | Comment / label / close | GitHub review comment |
+| **Validation** | AskUserQuestion before executing | AskUserQuestion before posting |
+
+**Decision rule**: use `/issue-triage` for issue backlog management, `/pr-triage` for code review backlog.
diff --git a/examples/skills/issue-triage/templates/issue-comment.md b/examples/skills/issue-triage/templates/issue-comment.md
new file mode 100644
index 0000000..c32af13
--- /dev/null
+++ b/examples/skills/issue-triage/templates/issue-comment.md
@@ -0,0 +1,107 @@
+# Issue Comment Templates
+
+Use these templates to generate GitHub issue comments during `/issue-triage` Phase 3. Comments are posted in **English** (international audience).
+
+---
+
+## Template 1 — Ack / Info Request
+
+Use when: issue is `Unclear` or `needs-info`, body is missing context, reproduction steps are absent.
+
+```markdown
+Thanks for opening this issue!
+
+To help us investigate, could you provide the following?
+
+- **Steps to reproduce**: A minimal sequence of actions that triggers the behavior
+- **Expected behavior**: What you expected to happen
+- **Actual behavior**: What actually happened
+- **Environment**: OS, version, relevant config (if applicable)
+
+Once we have this information, we can prioritize and route the issue appropriately.
+
+---
+*Triaged via Claude Code /issue-triage*
+```
+
+---
+
+## Template 2 — Duplicate
+
+Use when: Jaccard similarity >= 60% with an existing open or recently closed issue.
+
+```markdown
+Thanks for reporting this! After reviewing the backlog, this appears to be a duplicate of #{original_number}.
+
+{original_number} is tracking the same underlying behavior: {one-sentence description of original}.
+
+I'm closing this issue to consolidate discussion there. If you believe this is a distinct issue with different root cause or scope, please reopen with additional context explaining the difference.
+
+---
+*Triaged via Claude Code /issue-triage*
+```
+
+---
+
+## Template 3 — Close Stale
+
+Use when: issue has had no activity for >90 days and no assignee, or no response to a previous info request for >30 days.
+
+```markdown
+This issue has been inactive for {days} days without updates or response.
+
+We're closing it to keep the backlog actionable. If this is still relevant to you, please reopen and provide:
+
+- Current status: is this still reproducible?
+- Any additional context or workarounds you've found
+
+We're happy to pick this back up if it's still blocking you.
+
+---
+*Triaged via Claude Code /issue-triage*
+```
+
+---
+
+## Template 4 — Close Out of Scope
+
+Use when: issue describes functionality clearly outside the project's stated scope.
+
+```markdown
+Thanks for the suggestion! After review, this falls outside the current scope of this project.
+
+{Briefly explain why — e.g., "This project focuses on X; Y is handled by Z" or "This would require changes to the underlying architecture that are not planned."}
+
+You might find what you're looking for in:
+- {alternative project or tool if known}
+- {documentation link if relevant}
+
+Feel free to open a discussion if you'd like to explore this further.
+
+---
+*Triaged via Claude Code /issue-triage*
+```
+
+---
+
+## Formatting Rules
+
+**Tone**: Professional, direct, respectful. The reporter invested time to file the issue.
+
+**Rules**:
+- Never imply the reporter is wrong or wasted time
+- Be specific when referencing duplicates (title + number, not just number)
+- For close stale: always invite to reopen — do not make it feel permanent
+- For OOS: offer an alternative when possible, even if vague
+- No superlatives ("great issue", "awesome report") — factual only
+
+**Customization points** (replace in all templates):
+- `{original_number}`: issue number of the canonical duplicate
+- `{one-sentence description}`: short summary of the original issue
+- `{days}`: days since last activity (from `updatedAt`)
+- Inline explanations marked with `{...}`: always fill before posting
+
+**Signature** (required on all comments):
+```
+*Triaged via Claude Code /issue-triage*
+```
diff --git a/guide/cheatsheet.md b/guide/cheatsheet.md
index 36d7e69..571fad0 100644
--- a/guide/cheatsheet.md
+++ b/guide/cheatsheet.md
@@ -12,7 +12,7 @@ tags: [cheatsheet, reference]
 
 **Written with**: Claude (Anthropic)
 
-**Version**: 3.30.1 | **Last Updated**: February 2026
+**Version**: 3.30.2 | **Last Updated**: February 2026
 
 ---
 
@@ -613,4 +613,4 @@ where.exe claude; claude doctor; claude mcp list
 
 **Author**: Florian BRUNIAUX | [@Méthode Aristote](https://methode-aristote.fr) | Written with Claude
 
-*Last updated: February 2026 | Version 3.30.1*
+*Last updated: February 2026 | Version 3.30.2*
diff --git a/guide/ultimate-guide.md b/guide/ultimate-guide.md
index 3e0acc2..aec2cab 100644
--- a/guide/ultimate-guide.md
+++ b/guide/ultimate-guide.md
@@ -16,7 +16,7 @@ tags: [guide, reference, workflows, agents, hooks, mcp, security]
 
 **Last updated**: January 2026
 
-**Version**: 3.30.1
+**Version**: 3.30.2
 
 ---
 
@@ -4946,7 +4946,7 @@ The `.claude/` folder is your project's Claude Code directory for memory, settin
 | Personal preferences | `CLAUDE.md` | ❌ Gitignore |
 | Personal permissions | `settings.local.json` | ❌ Gitignore |
 
-### 3.30.1 Version Control & Backup
+### 3.30.2 Version Control & Backup
 
 **Problem**: Without version control, losing your Claude Code configuration means hours of manual reconfiguration across agents, skills, hooks, and MCP servers.
 
@@ -22508,4 +22508,4 @@ We'll evaluate and add it to this section if it meets quality criteria.
 
 **Contributions**: Issues and PRs welcome.
 
-**Last updated**: January 2026 | **Version**: 3.30.1
+**Last updated**: January 2026 | **Version**: 3.30.2
diff --git a/machine-readable/reference.yaml b/machine-readable/reference.yaml
index 34ceedc..c98d58d 100644
--- a/machine-readable/reference.yaml
+++ b/machine-readable/reference.yaml
@@ -3,7 +3,7 @@
 # Source: guide/ultimate-guide.md
 # Purpose: Condensed index for LLMs to quickly answer user questions about Claude Code
 
-version: "3.30.1"
+version: "3.30.2"
 updated: "2026-03-03"
 
 # ════════════════════════════════════════════════════════════════
@@ -883,6 +883,14 @@ deep_dive:
   pr_triage_phases: "3 (audit + deep review + validated comments)"
   pr_triage_when: "5+ open PRs, periodic backlog triage, post-sprint cleanup"
   pr_triage_vs_review_pr: "pr-triage = backlog N PRs | review-pr = single PR focus"
+  # Issue Triage Skill (added 2026-03-05)
+  issue_triage_skill: "examples/skills/issue-triage/SKILL.md"
+  issue_triage_template: "examples/skills/issue-triage/templates/issue-comment.md"
+  issue_triage_phases: "3 (audit + deep analysis + validated actions)"
+  issue_triage_when: "10+ open issues, periodic backlog cleanup, duplicate detection"
+  issue_triage_jaccard: "Jaccard similarity >= 0.60 on normalized title+body tokens — self-contained, no external lib"
+  issue_triage_thresholds: "staleness_days=30, very_stale_days=90, jaccard_threshold=60%, closed_compare=20, open_limit=100"
+  issue_triage_vs_pr_triage: "issue-triage = issue backlog (comment/label/close) | pr-triage = PR backlog (code review comment)"
   # Cowork documentation (v1.0 - migrated to dedicated repo)
   cowork_reference: "machine-readable/cowork-reference.yaml"  # Dedicated YAML index (kept local)
   cowork_hub: "https://github.com/FlorianBruniaux/claude-cowork-guide/blob/main/README.md"
@@ -1438,7 +1446,7 @@ ecosystem:
       - "Cross-links modified → Update all 4 repos"
     history:
       - date: "2026-01-20"
-        event: "Code Landing sync v3.30.1, 66 templates, cross-links"
+        event: "Code Landing sync v3.30.2, 66 templates, cross-links"
         commit: "5b5ce62"
       - date: "2026-01-20"
         event: "Cowork Landing fix (paths, README, UI badges)"
@@ -1450,7 +1458,7 @@ ecosystem:
 onboarding_matrix_meta:
   version: "2.0.0"
   last_updated: "2026-02-05"
-  aligned_with_guide: "3.30.1"
+  aligned_with_guide: "3.30.2"
   changelog:
     - version: "2.0.0"
       date: "2026-02-05"
@@ -1478,7 +1486,7 @@ onboarding_matrix:
       core: [rules, sandbox_native_guide, commands]
       time_budget: "5 min"
       topics_max: 3
-      note: "SECURITY FIRST - sandbox before commands (v3.30.1 critical fix)"
+      note: "SECURITY FIRST - sandbox before commands (v3.30.2 critical fix)"
 
     beginner_15min:
       core: [rules, sandbox_native_guide, workflow, essential_commands]
@@ -1563,7 +1571,7 @@ onboarding_matrix:
         - default: agent_validation_checklist
       time_budget: "60 min"
       topics_max: 6
-      note: "Dual-instance pattern for quality workflows (v3.30.1)"
+      note: "Dual-instance pattern for quality workflows (v3.30.2)"
 
   learn_security:
     intermediate_30min:
@@ -1574,7 +1582,7 @@ onboarding_matrix:
         - default: permission_modes
       time_budget: "30 min"
       topics_max: 4
-      note: "NEW goal (v3.30.1) - Security-focused learning path"
+      note: "NEW goal (v3.30.2) - Security-focused learning path"
 
     power_60min:
       core: [sandbox_native_guide, mcp_secrets_management, security_hardening]
@@ -1599,7 +1607,7 @@ onboarding_matrix:
       core: [rules, sandbox_native_guide, workflow, essential_commands, context_management, plan_mode]
       time_budget: "60 min"
       topics_max: 6
-      note: "Security foundation + core workflow (v3.30.1 sandbox added)"
+      note: "Security foundation + core workflow (v3.30.2 sandbox added)"
 
     intermediate_120min:
       core: [plan_mode, agents, skills, config_hierarchy, git_mcp_guide, hooks, mcp_servers]