Vault / index.md

Index

Catalog of every note in the vault. Update on every add.

Two ways in:

Take the course (guided, from zero): README → catalog → README. Run the repo in Claude Code/Cowork and CLAUDE tutors you and tracks progress in learner-db.json.
Use as reference (cert prep): cca-f-exam-overview → study-plan → cheatsheet.

🎓 The Course (mirrors every Anthropic course)

claude-overview — what Claude is and the surfaces it ships on
model-family — Opus / Sonnet / Haiku, picking a model, current IDs
claude-ai-features — Projects, Artifacts, file uploads, MCP connectors
prompt-engineering-basics — the 5+2 prompt levers and common failures
anthropic-safety — Constitutional AI, HHH, Responsible Scaling Policy
claude-cowork — Claude working directly on files, folders, and apps; skills, plugins, standing context
ai-fluency-4d — Anthropic's 4D Framework: Delegation, Description, Discernment, Diligence
ai-capabilities-limitations — four machine properties: prediction, knowledge, working memory, steerability

messages-api — request/response shape, content blocks, stop_reason
system-prompts — structure, principles vs conditionals, dilution
tool-use — the tool loop, schemas, tool_choice, common pitfalls
parallel-tool-use — multiple tool_use in one turn; latency wins
computer-use — beta desktop control: screenshot → click/type loop, versions, safety
streaming — SSE events, streaming + tools + thinking
prompt-caching — cache_control, breakpoints, TTLs, CALM
extended-thinking — when to enable thinking; signatures
vision — image and PDF input
multi-turn-conversations — statelessness, turn rules, compaction
structured-output — three techniques ranked, schemas, validation
few-shot-prompting — choosing examples, XML tags, classification
rag-patterns — long-context vs RAG, contextual retrieval, citations
batch-api — bulk offline jobs at 50% off
claude-bedrock — Claude on AWS Bedrock, diffs from first-party
context-management — compaction strategies, state objects, token budgets

claude-code-overview — surfaces, built-ins, permission model
claude-code-md — CLAUDE.md hierarchy and what belongs where
claude-code-settings — settings.json, permissions, output formats
claude-code-slash-commands — .claude/commands/*.md and $ARGUMENTS
claude-code-hooks — lifecycle hooks for deterministic guardrails
claude-code-plan-mode — propose-then-execute for ambiguous tasks
agent-skills — reusable capability bundles via SKILL.md
subagents — isolated delegate agents
agent-sdk — the library Claude Code is built on
agentic-patterns — the five canonical agent shapes
claude-platform-101 — Developer Platform intro: agent loop, managed vs self-hosted agents, costs

certifications-overview — all four exams: Pearson VUE, scoring, retakes, renewal, partner discounts
ccao-f-exam — Associate – Foundations: 7 domains for Claude-as-productivity-tool users
ccdv-f-exam — Developer – Foundations: 8 domains, Applications & Integration at 33.1%
ccar-p-exam — Architect – Professional: 7 domains, governance and stakeholder work at 28%

storm-overview — Stanford's multi-perspective research pipeline
storm-pipeline — the four STORM modules, stage by stage
storm-for-exams — QSTORM: adapting STORM to exam-question generation
co-storm — human-in-the-loop discovery of unknown unknowns
co-storm-for-notes — Co-STORM retargeted to research broad new vault notes (the deep-note path)
reverse-prompting-overview — flipped interaction + reverse prompt engineering
reverse-prompting-patterns — quiz-me loops, elicitation wizards, structured branching

2026-05-28-karpathy-vault-pattern — the three-layer vault pattern
2026-05-28-cca-f-overview — CCA-F exam details and learning-path inventory
2026-05-28-building-effective-agents — Anthropic's canonical agentic patterns post
2026-05-28-mcp-architecture — official MCP architecture docs
2026-05-28-claude-code-overview — official Claude Code overview (code.claude.com)
2026-05-28-prompt-engineering-index — pointer to "Prompting best practices"