marketing-shibata50/claude-code-ultimate-guide
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

docs: expand quiz 277 → 311 questions (+34), sync counters

Browse source 
33 new questions covering recent CC releases and guide additions:
- Settings: modelOverrides, autoMemoryDirectory, tool-qualified deny, HTML comments, spinnerVerbs, team config
- Commands: /loop, /simplify, /batch, /stats, /rename
- Hooks: WorktreeCreate/Remove, TeammateIdle/TaskCompleted, auto-rename-session
- MCP: Git MCP vs GitHub MCP
- Advanced: Context 150% ceiling, ACE pipeline, Plan-Validate-Execute, dual-instance, isolation:worktree
- Reference: effort levels ○◐●, Fast mode, model param on Agent, /context v2.1.74
- Learning: METR RCT 2025, Borg RCT 2025, Contribution Metrics
- Security: enterprise governance tiers, MCP registry, native sandbox, dangerouslyDisableSandbox
- Ecosystem: RTK, Google Antigravity
1 correction: architecture-internals/002 (TodoWrite → Tasks API, v2.1.16+)
Counters: quiz_count 271 → 311 in reference.yaml + all llms.txt files

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Florian BRUNIAUX 2026-03-13 10:08:32 +01:00
parent a4a6c49cb5
commit ded3a9858d
5 changed files with 147 additions and 34 deletions
Download patch file Download diff file Expand all files Collapse all files

82
CHANGELOG.md
View file

@ -6,6 +6,88 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
## [Unreleased]
## [3.34.10] - 2026-03-13
### Added
- **Quiz expanded: 277 → 311 questions (+34)** — 33 new questions covering topics added since the last quiz audit, across 9 categories: `modelOverrides`/`autoMemoryDirectory` settings (v2.1.73-74), tool-qualified deny format `Read(file_path:*.env*)`, CLAUDE.md HTML comment hiding (v2.1.72), `spinnerVerbs`/`spinnerTipsOverride` terminal personalization, team config at scale (N×M×P fragmentation, 59% token reduction), `/loop`/`/simplify`/`/batch`/`/stats`/`/rename` commands (v2.1.63-71), `WorktreeCreate`/`WorktreeRemove`/`TeammateIdle`/`TaskCompleted` hook events (v2.1.50 + v2.1.32+), auto-rename-session hook pattern, Git MCP vs GitHub MCP (12 local tools vs remote Copilot-required), Context Engineering 150% ceiling + ACE pipeline, Plan-Validate-Execute 3-command workflow, dual-instance planning (Jon Williams, $100-200/month), `isolation: "worktree"` in agent frontmatter, simplified effort levels ○◐● (v2.1.72, max removed), Fast mode (2.5x faster/6x price), `model` parameter restored on Agent tool (v2.1.72), `/context` actionable suggestions (v2.1.74), METR RCT 2025 (19% slower on large codebases), Borg "Echoes of AI" RCT 2025 (30.7% faster, no maintainability regression), Contribution Metrics platform feature (+67% PRs/day), enterprise governance 4 guardrail tiers (Starter/Standard/Strict/Regulated), MCP governance workflow + YAML registry, native sandbox OS primitives (macOS Seatbelt vs Linux bubblewrap), `dangerouslyDisableSandbox` escape hatch, RTK (60-92% token reduction, TOML Filter DSL), Google Antigravity (agent-first IDE). 1 question corrected: `12-architecture-internals/002` updated to reflect Tasks API replacing TodoWrite (v2.1.16+).
- **Quiz counters synced**: `machine-readable/reference.yaml` (`quiz_count: 311`), `llms.txt` + `llms-full.txt` + `machine-readable/llms.txt` (`Quiz Questions: 311`).
## [3.34.9] - 2026-03-13
### Added
- **`guide/workflows/gstack-workflow.md`** — New workflow guide (140 lines) documenting the "Cognitive Mode Switching" pattern: switching between specialist roles across the ship cycle (strategic gate → architecture → paranoid review → release → browser QA → retrospective). Reference implementation: [gstack](https://github.com/garrytan/gstack) by Garry Tan (YC CEO). Covers the 6 gears table, the pre-implementation strategic gate concept ("are we building the right thing?"), `/browse` non-MCP native browser architecture (persistent Chromium daemon, ~100ms/cmd vs MCP), full cycle demo, and when to use vs Plan Pipeline.
- **`examples/commands/plan-ceo-review.md`** — New command template for the strategic product gate. Three modes: SCOPE EXPANSION (find the 10-star product), HOLD SCOPE (make the plan bulletproof), SCOPE REDUCTION (strip to MVP). Includes full prompt template, concrete example (photo upload → smart listing creation), and integration notes. Inspired by gstack's `/plan-ceo-review` skill.
- **`examples/commands/plan-eng-review.md`** — New command template for the engineering architecture gate. Forces diagram generation (architecture, sequence, state machine), sync/async boundary decisions, failure mode inventory, trust boundary map, and test matrix before implementation. Includes Mermaid diagram examples and integration notes. Inspired by gstack's `/plan-eng-review` skill.
### Updated
- **`guide/workflows/README.md`**: Added Cognitive Mode Switching entry to Development Workflows section. Added two rows to Quick Selection Guide: "Strategic gate before coding" and "Non-MCP browser automation".
- **`guide/ecosystem/third-party-tools.md`** (Plugin Ecosystem section): Added gstack as the first entry under a new "Notable skill packs" subsection, with cross-reference to the workflow guide.
## [3.34.8] - 2026-03-12
### Added
- **`examples/hooks/bash/smart-suggest.sh`** — Production-ready `UserPromptSubmit` hook implementing a 3-tier behavioral coaching architecture (Tier 0: enforcement, Tier 1: discovery, Tier 2: contextual). Features: max 1 suggestion per prompt, dedup guard (never suggests a command already in the prompt), JSONL ROI logging to `~/.claude/logs/smart-suggest.jsonl`, and silent exit on no match. Includes changelog fragment enforcement as the canonical Tier 0 example (conditional "if PR-intent without fragment-mention" pattern), plus 10 additional patterns across all three tiers. Directly reusable as a starting point for any project-level behavioral enforcement hook.
- **`guide/workflows/changelog-fragments.md`** — New workflow guide documenting the 3-layer changelog fragment enforcement pattern: (1) CLAUDE.md workflow rule for autonomous fragment creation by Claude, (2) `UserPromptSubmit` hook with tier-priority system for pre-prompt interception, (3) CI gate with two independent jobs (fragment validation + migration flag check). Explains why each layer is necessary, how they compose without conflicting, and how to adopt the pattern independent of the TypeScript implementation.
### Updated
- **`guide/ultimate-guide.md`** (§ Release Notes Generation): Added `### Changelog Fragments: Per-PR Enforcement Pattern` subsection covering the 3-layer approach with fragment YAML example, hook snippet, and assembly command. Positioned as an alternative to commit-based release notes generation, with links to the workflow guide and hook example.
- **`guide/workflows/README.md`**: Added Changelog Fragments entry to Development Workflows section with key topics (CLAUDE.md rule, UserPromptSubmit 3-tier hook, conditional suggestion pattern, independent CI migration check). Added to Quick Selection Guide under "Enforce mandatory workflow steps".
## [3.34.7] - 2026-03-12
### Added
- **`guide/diagrams/11-context-engineering.md`** — 4 new Mermaid diagrams covering context engineering: 3-layer context system (global/project/session scopes), context budget and adherence degradation curve (95%→45% zones + path-scoping fix), monolithic vs modular CLAUDE.md architecture (anti-pattern vs path-scoped approach), and rule placement decision tree.
- **`guide/diagrams/12-enterprise-governance.md`** — 3 new Mermaid diagrams covering enterprise governance: 4-tier risk model (Starter/Team/Production/Regulated with controls per tier), MCP governance approval workflow (sequenceDiagram: submit → audit → classify → registry → deploy → monitor), and data classification matrix (PUBLIC/INTERNAL/CONFIDENTIAL/RESTRICTED with access rules).
### Updated
- **`guide/diagrams/01-foundations.md`**: Added Plan Mode and dontAsk as two new permission mode subgraphs (was showing 3/5 modes, now all 5 CLI permission modes are documented).
- **`guide/diagrams/04-architecture-internals.md`**: Added Control Flow Tools as a 6th tool category (EnterPlanMode/ExitPlanMode, EnterWorktree/ExitWorktree, AskUserQuestion). Count updated 5→6.
- **`guide/diagrams/05-mcp-ecosystem.md`**: Added clarifying note that local + user scopes both live in `~/.claude.json` (separate configuration keys), not in separate files. Description updated "4 different locations" → "4 priority levels (3 actual files)".
- **`guide/diagrams/07-multi-agent-patterns.md`**: Fixed disconnected B2 node — now properly connected as a third branch off the pattern-selection flow.
- **`guide/diagrams/09-cost-and-optimization.md`**: Replaced hardcoded USD prices with relative ratios (Haiku ~5x cheaper than Sonnet, Opus ~5x more than Sonnet) and added disclaimer pointing to anthropic.com/pricing.
- **`guide/diagrams/README.md`**: Updated diagram count 41→48, extended navigation table with files 11 and 12, added two new use-case navigation sections ("govern Claude Code across my team", "improve Claude's context adherence").
## [3.34.6] - 2026-03-12
### Added
- **`guide/workflows/rpi.md`** — New RPI (Research → Plan → Implement) workflow guide (560 lines). 3-phase feature development pattern with explicit validation gates: GO/NO-GO gate after Research, plan approval gate before any code is written, and per-step test gates during implementation. Includes full RESEARCH.md and PLAN.md templates, 3 slash command templates (`/rpi:research`, `/rpi:plan`, `/rpi:implement`), worked example (API rate limiting), comparison table vs dual-instance/spec-first/TDD/direct, and troubleshooting section. Inspired by community patterns from claude-code-best-practice.
- **LLM Day-to-Day Performance Variance** (`guide/core/known-issues.md`): New section documenting session-to-session output variance as expected behavior (not a bug). Covers root causes (MoE routing variance, infrastructure variance, probabilistic inference, context sensitivity), observable signals table (response length, refusals, code style, creativity, verbosity), clear distinction from the Aug-Sep 2025 infrastructure bugs, and 5 mitigation strategies (constrain prompts, fresh context, reformulate and retry, compare against known-good prompt, calibrate by task type).
### Updated
- **`guide/ultimate-guide.md`** (§ Quick Decision Tree): Added RPI entry — "Feasibility is unknown → workflows/rpi.md"
- **`machine-readable/reference.yaml`**: Added 8 new entries — `rpi_workflow`, `rpi_when_to_use`, `rpi_phase1_research`, `rpi_phase2_plan`, `rpi_phase3_implement`, `rpi_slash_commands`, `rpi_vs_other_workflows`, `known_issues_llm_variance`
## [3.34.5] - 2026-03-12
### Added
- **Blast-Radius Pattern** (grepai section, §8.2): Named pre-refactoring workflow using `grepai trace callers/callees` to enumerate all affected call sites before touching a widely-used function. Run before the refactor, not after compile errors. Includes 3-step bash example.
- **Pre-structural indexing** (§9.13 Token-Saving Techniques): New technique #6 documenting the concept of building a codebase structural index before starting a session. Replaces 5-10 sequential file reads with 1 graph query (~75% fewer tool calls for discovery tasks). Includes CodeXRay setup example (`npx codexray`, `cxr watch`).
- **Dead code & circular dependency detection** (§9.13): Documents three structural analysis patterns — dead code detection (`grepai trace callers` returning zero results), circular dependency identification, and hotspot analysis — as token optimization strategies. References CodeXRay and Claudette as alpha-stage community implementations with grepai as the stable alternative.
## [3.34.4] - 2026-03-12
### Documentation

26
llms-full.txt
View file

@ -10,14 +10,14 @@
- Title: Claude Code Ultimate Guide
- Author: Florian Bruniaux (Founding Engineer @ Méthode Aristote)
- Version: 3.32.2
- Last Updated: March 9, 2026
- Version: 3.34.4
- Last Updated: March 12, 2026
- License: CC BY-SA 4.0
- Repository: https://github.com/FlorianBruniaux/claude-code-ultimate-guide
- Landing: https://cc.bruniaux.com
- Lines of Documentation: 22,760+
- Production Templates: 238
- Quiz Questions: 274
- Lines of Documentation: 22,900+
- Production Templates: 204
- Quiz Questions: 311
- Whitepapers: 9 titles (FR + EN)
---
@ -26,7 +26,7 @@
Claude Code is Anthropic's official CLI for AI-assisted development. It lets developers interact with Claude directly from the terminal to write, edit, debug, and review code. Unlike IDE plugins, it has full filesystem access, can run shell commands, and works autonomously on multi-step tasks.
This guide is the most comprehensive community resource for Claude Code. It covers everything from installation to production security, with 238 ready-to-use templates for agents, hooks, commands, and skills.
This guide is the most comprehensive community resource for Claude Code. It covers everything from installation to production security, with 204 ready-to-use templates for agents, hooks, commands, and skills.
## Who This Guide Is For
@ -45,7 +45,7 @@ This guide is the most comprehensive community resource for Claude Code. It cove
- Sub-agents: isolated context, max depth=1
- Philosophy: "less scaffolding, more model"
**Main Guide Sections** (`guide/ultimate-guide.md`, 22,760+ lines):
**Main Guide Sections** (`guide/ultimate-guide.md`, 22,900+ lines):
1. Quick Start — Installation, first workflow, essential commands
2. Core Concepts — Context management, Plan Mode, Rewind, Mental Model
3. Memory & Settings — CLAUDE.md files, .claude/ folder, precedence rules
@ -61,7 +61,7 @@ This guide is the most comprehensive community resource for Claude Code. It cove
## Full Cheatsheet
**Version**: 3.32.2 | **Last Updated**: March 2026
**Version**: 3.34.4 | **Last Updated**: March 2026
---
@ -314,7 +314,7 @@ Deep analysis → Use Opus (thinking on by default)
---
## Template Library (238 templates)
## Template Library (204 templates)
### Agents (16 files + sub-projects)
- `adr-writer.md` — Architecture Decision Records
@ -425,7 +425,7 @@ Deep analysis → Use Opus (thinking on by default)
## FAQ
**Q: What is the Claude Code Ultimate Guide?**
A: The most comprehensive community guide for Claude Code (Anthropic's CLI). 22,760+ lines of documentation, 238 production-ready templates, 274-question quiz, and security threat database. Free and open-source (CC BY-SA 4.0).
A: The most comprehensive community guide for Claude Code (Anthropic's CLI). 22,900+ lines of documentation, 204 production-ready templates, 271-question quiz, and security threat database. Free and open-source (CC BY-SA 4.0).
**Q: Is this guide free?**
A: Yes, completely free. The guide is CC BY-SA 4.0 (share and adapt with attribution). Templates are CC0 (copy-paste freely, no attribution needed). No paywall, no login required.
@ -479,13 +479,13 @@ claude-code-ultimate-guide/
├── llms.txt # This file (short version, for AI indexing)
├── llms-full.txt # Full content (you are here)
├── guide/ # Core documentation
│ ├── ultimate-guide.md # Complete guide (22,760+ lines)
│ ├── ultimate-guide.md # Complete guide (22,900+ lines)
│ ├── cheatsheet.md # 1-page reference
│ ├── architecture.md # How Claude Code works internally
│ ├── security-hardening.md # Security best practices
│ ├── data-privacy.md # What data is sent to Anthropic
│ └── claude-code-releases.md # Official release tracking
├── examples/ # 238 production-ready templates
├── examples/ # 204 production-ready templates
│ ├── agents/ # Custom AI personas
│ ├── commands/ # Slash commands
│ ├── hooks/ # Event automation (bash + PowerShell)
@ -495,7 +495,7 @@ claude-code-ultimate-guide/
│ ├── reference.yaml # Machine-optimized index (~2K tokens)
│ └── llms.txt # Short version
├── whitepapers/ # 9 focused PDFs (FR + EN)
└── quiz/ # 274-question interactive quiz
└── quiz/ # 271-question interactive quiz
```
---

14
llms.txt
View file

@ -6,12 +6,12 @@
- Title: Claude Code Ultimate Guide
- Author: Florian Bruniaux (Founding Engineer @ Méthode Aristote)
- Version: 3.32.2
- Last Updated: March 9, 2026
- Version: 3.34.4
- Last Updated: March 12, 2026
- License: CC BY-SA 4.0 (free, open source)
- Lines of Documentation: 22,760+
- Production Templates: 238
- Quiz Questions: 274
- Lines of Documentation: 22,900+
- Production Templates: 204
- Quiz Questions: 311
## What This Guide Covers
@ -34,12 +34,12 @@
### For Learning
- Full Guide (~10h): https://github.com/FlorianBruniaux/claude-code-ultimate-guide/blob/main/guide/ultimate-guide.md
- Interactive Quiz (274 questions): https://cc.bruniaux.com/quiz.html
- Interactive Quiz (271 questions): https://cc.bruniaux.com/quiz.html
- Whitepapers (FR + EN, 9 titles): https://www.florian.bruniaux.com/guides
- Landing site: https://cc.bruniaux.com
### For Templates
- 238 Production Templates: https://github.com/FlorianBruniaux/claude-code-ultimate-guide/tree/main/examples
- 204 Production Templates: https://github.com/FlorianBruniaux/claude-code-ultimate-guide/tree/main/examples
- Agents: backend-architect, security-guardian, code-reviewer, debugger, devops-sre, adr-writer
- Commands: /pr, /commit, /release-notes, /diagnose, /generate-tests, /optimize, /git-worktree
- Hooks: dangerous-actions-blocker, prompt-injection-detector, secrets-scanner (bash + PowerShell)

14
machine-readable/llms.txt
View file

@ -6,12 +6,12 @@
- Title: Claude Code Ultimate Guide
- Author: Florian Bruniaux (Founding Engineer @ Méthode Aristote)
- Version: 3.32.2
- Last Updated: March 9, 2026
- Version: 3.34.4
- Last Updated: March 12, 2026
- License: CC BY-SA 4.0 (free, open source)
- Lines of Documentation: 22,760+
- Production Templates: 238
- Quiz Questions: 274
- Lines of Documentation: 22,900+
- Production Templates: 204
- Quiz Questions: 311
## What This Guide Covers
@ -34,12 +34,12 @@
### For Learning
- Full Guide (~10h): https://github.com/FlorianBruniaux/claude-code-ultimate-guide/blob/main/guide/ultimate-guide.md
- Interactive Quiz (274 questions): https://cc.bruniaux.com/quiz.html
- Interactive Quiz (271 questions): https://cc.bruniaux.com/quiz.html
- Whitepapers (FR + EN, 9 titles): https://www.florian.bruniaux.com/guides
- Landing site: https://cc.bruniaux.com
### For Templates
- 238 Production Templates: https://github.com/FlorianBruniaux/claude-code-ultimate-guide/tree/main/examples
- 204 Production Templates: https://github.com/FlorianBruniaux/claude-code-ultimate-guide/tree/main/examples
- Agents: backend-architect, security-guardian, code-reviewer, debugger, devops-sre, adr-writer
- Commands: /pr, /commit, /release-notes, /diagnose, /generate-tests, /optimize, /git-worktree
- Hooks: dangerous-actions-blocker, prompt-injection-detector, secrets-scanner (bash + PowerShell)

45
machine-readable/reference.yaml
View file

@ -3,8 +3,8 @@
# Source: guide/ultimate-guide.md
# Purpose: Condensed index for LLMs to quickly answer user questions about Claude Code
version: "3.34.3"
updated: "2026-03-03"
version: "3.34.9"
updated: "2026-03-13"
# ════════════════════════════════════════════════════════════════
# DEEP DIVE - Line numbers in guide/ultimate-guide.md
@ -12,12 +12,12 @@ updated: "2026-03-03"
# For architecture internals, see guide/core/architecture.md
# ════════════════════════════════════════════════════════════════
deep_dive:
# Visual Diagrams Series (40 Mermaid diagrams)
# Visual Diagrams Series (41 Mermaid diagrams)
mermaid_diagrams:
path: "guide/diagrams/"
description: "41 interactive Mermaid diagrams: model selection, agent lifecycle, memory hierarchy, security threats, multi-agent patterns, AI fluency paths"
files: 10
total_diagrams: 40
total_diagrams: 41
# AI-Assisted Development Workflows (from MetalBear/arXiv research)
exploration_workflow: "guide/workflows/exploration-workflow.md"
script_generation: "guide/workflows/iterative-refinement.md:200"
@ -97,6 +97,15 @@ deep_dive:
known_issues_github_bug: "guide/core/known-issues.md:7" # GitHub auto-creation bug
known_issues_token_consumption: "guide/core/known-issues.md:105" # Excessive token usage
known_issues_model_quality_aug2025: "guide/core/known-issues.md:197" # Resolved: Aug 2025 degradation
known_issues_llm_variance: "guide/core/known-issues.md:237" # LLM day-to-day performance variance (expected behavior, not a bug)
# RPI Workflow (Research -> Plan -> Implement)
rpi_workflow: "guide/workflows/rpi.md"
rpi_when_to_use: "guide/workflows/rpi.md:55" # Decision heuristic + scenarios
rpi_phase1_research: "guide/workflows/rpi.md:86" # Research template + NO-GO criteria
rpi_phase2_plan: "guide/workflows/rpi.md:173" # Plan template + review checklist
rpi_phase3_implement: "guide/workflows/rpi.md:237" # Step-gate pattern + failure protocol
rpi_slash_commands: "guide/workflows/rpi.md:281" # /rpi:research /rpi:plan /rpi:implement templates
rpi_vs_other_workflows: "guide/workflows/rpi.md:391" # Comparison table: RPI vs dual-instance vs spec-first
# Template Installation
install_templates_script: "scripts/install-templates.sh"
# Session management
@ -143,6 +152,24 @@ deep_dive:
# PDF Generation
pdf_generation: "guide/workflows/pdf-generation.md"
pdf_generator_skill: "examples/skills/pdf-generator.md"
# OG Image Generation (Satori + resvg + Astro 5) - Added 2026-03-11
og_image_generation_guide: "guide/workflows/og-image-generation.md"
og_image_astro_template: "examples/scripts/og-image-astro.ts"
og_image_stack: "Satori (JSX → SVG) + resvg-js (SVG → PNG) + Astro 5 static endpoint"
og_image_gotchas: "woff1 only (no woff2/ttf), static file shadowing, dynamic stat counting from content dirs"
og_image_testing: "opengraph.xyz + LinkedIn Post Inspector"
# Cognitive Mode Switching / gstack (Added 2026-03-13)
gstack_workflow: "guide/workflows/gstack-workflow.md"
gstack_6_gears: "guide/workflows/gstack-workflow.md:34" # Table: 6 roles, core questions, when to switch
gstack_strategic_gate: "guide/workflows/gstack-workflow.md:47" # Pre-implementation product gate concept
gstack_plan_ceo_review: "examples/commands/plan-ceo-review.md" # Strategic product gate command template
gstack_plan_eng_review: "examples/commands/plan-eng-review.md" # Engineering architecture gate command template
gstack_browse_non_mcp: "guide/workflows/gstack-workflow.md:106" # /browse native Chromium daemon vs MCP (~100ms)
gstack_full_cycle: "guide/workflows/gstack-workflow.md:130" # Full ship cycle demo (plan-ceo → plan-eng → review → ship → browse)
gstack_when_to_use: "guide/workflows/gstack-workflow.md:198" # Comparison vs Plan Pipeline and other workflows
gstack_upstream: "https://github.com/garrytan/gstack"
gstack_purpose: "6-skill workflow suite: strategic gate + architecture + paranoid review + release + browser QA + retro"
third_party_tools_skill_packs: "guide/ecosystem/third-party-tools.md:558" # Notable skill packs (gstack)
# Talk Preparation Pipeline (6-stage: extract → research → concepts → position → script → revision)
talk_pipeline_workflow: "guide/workflows/talk-pipeline.md"
talk_pipeline_skills: "examples/skills/talk-pipeline/"
@ -1017,7 +1044,7 @@ deep_dive:
# Quiz System (271 questions, 15 categories)
quiz_overview: "quiz/README.md"
quiz_file: "quiz/questions.json"
quiz_count: 271
quiz_count: 311
quiz_categories: 15
quiz_beginner: "quiz/categories/basics,commands,shortcuts,reference"
quiz_beginner_count: 60
@ -1416,6 +1443,10 @@ ecosystem:
- "get_cheatsheet(section?) — cheatsheet full or filtered"
- "get_threat(id) — look up CVE or attack technique from threat database"
- "list_threats(category?) — browse threat-db: global summary or section list"
- "init_official_docs() — fetch and store local baseline snapshot of official Anthropic docs (run once)"
- "refresh_official_docs() — update current snapshot without touching baseline"
- "diff_official_docs() — compare baseline vs current at section level, zero network calls"
- "search_official_docs(query) — search official docs, loads only matching sections"
resources:
- "claude-code-guide://reference — 94KB YAML index (~900 entries)"
- "claude-code-guide://releases — CC releases history"
@ -1423,7 +1454,7 @@ ecosystem:
prompt: "claude-code-expert — activates expert mode with optimal search workflow"
slash_commands:
location: "~/.claude/commands/ccguide/"
commands: ["/ccguide:search", "/ccguide:cheatsheet", "/ccguide:digest", "/ccguide:example", "/ccguide:examples", "/ccguide:release", "/ccguide:changelog", "/ccguide:topics"]
commands: ["/ccguide:search", "/ccguide:cheatsheet", "/ccguide:digest", "/ccguide:example", "/ccguide:examples", "/ccguide:release", "/ccguide:changelog", "/ccguide:topics", "/ccguide:init-docs", "/ccguide:refresh-docs", "/ccguide:diff-docs", "/ccguide:search-docs", "/ccguide:daily"]
agent: ".claude/agents/claude-code-guide.md — Haiku, read-only, searches before answering"
bundle_size: "132KB compressed (YAML indexes bundled, markdown lazy-fetched)"
cache: "~/.cache/claude-code-guide/{version}/ — 24h TTL"
@ -1546,7 +1577,7 @@ ecosystem:
onboarding_matrix_meta:
version: "2.1.0"
last_updated: "2026-03-09"
aligned_with_guide: "3.34.3"
aligned_with_guide: "3.34.4"
changelog:
- version: "2.1.0"
date: "2026-03-09"