Skip to content
CCAF
Study Site
Agentic Architecture & Orchestration
27%
Agentic loop
Stop reasons
pause_turn handling
Tool execution types (client vs server)
Subagent decision criteria
Adaptive thinking and effort
Interleaved thinking
Extended thinking with tool use
Task and Agent tool
AgentDefinition
Hub-and-spoke orchestration
Task decomposition (chaining vs adaptive)
State across context windows
Workflow tool
Workflows vs Agents
Augmented LLM
Prompt Chaining Workflow
Routing Workflow
Parallelization Workflow
Orchestrator-Workers Workflow
Evaluator-Optimizer Workflow
Autonomous Agents
Architect's Reference Matrix
Goal-oriented delegation
PreToolUse hook enforcement
Structured Intermediate Representations
The Production Architecture Blueprint
Zero-Tolerance Compliance
Claude Code Configuration & Workflows
20%
Config Hierarchy
CLAUDE.md Hierarchy
CLAUDE.md Imports And Rules
Skills And Slash Commands
Auto Memory
Permissions Merge And Deny Beats Allow
Hook Types And Events
Hook Exit Codes And JSON Output
Checkpointing and Rewind
Output Styles
Custom Slash Commands
Plan Mode And Explore Subagent
Filesystem Subagents (.claude Agents)
CI CD Headless Flags
Session Resume And Fork
Prompt Engineering & Structured Output
20%
Prompt Clarity And Motivation
Role Prompting
XML Structuring
Few-Shot Prompting
Long-Context Ordering
Prefill Migration
Structured Outputs
Validation-Retry Loops
Message Batches API
Multi-Pass And Multi-Instance Review
Opus Severity-Filter Trap
Building Evals
Prompt Technique Taxonomy
Mathematical consistency enforcement
Resilient schema design
Routing for cost and SLA
Tool Design & MCP Integration
18%
Tool Definition Schema
Tool Name Regex
Strict Tool Use
Input Examples
Effective Tool Descriptions
Tool Choice
MCP Connector
MCPToolset Allowlist and Denylist
Defer Loading and Tool Search
Structured Tool Errors
Subagent Error Handling
MCP Config Scopes
Built-in Tools
Tool Namespacing
Poka-Yoke Tool Design
Agentic Tool Evaluation Loop
Token-Efficient Tool Responses
MCP Primitives
MCP Architecture
Token-Efficient Tool Use
Context Management & Reliability
15%
Context Windows By Model
Context Rot
Lost In The Middle
Context Awareness
Server-side Compaction
Progressive Summarisation Risk
Scratchpad Pattern
Context Engineering
System Prompt Altitude
Just-in-Time Retrieval
Structured Note-Taking
Context Editing
Memory Tool
Prompt Caching
Caching Plus Batching
ZDR Eligibility
Model Lineup And Pricing
model_context_window_exceeded behaviour
Error Propagation In Multi-Agent Systems
Stratified Sampling And Confidence Scores
Tool Context Pruning
Escalation to Human
unscored
Escalation triggers
Wrong escalation signals
Explicit-request-overrides rule
Structured handoff summary
Minimal footprint principle
Human-in-the-loop calibration
Limits of automated retry
The Escalation Handoff
Claude's Constitution
Architect's Playbook Patterns
cross-cutting
Compressing Long Sessions
Forcing Execution Order
Graceful Tool Failure
MCP Tool Specificity
Parallelization and Caching
Shared Memory Architecture
Today
Tutor
Ready
Map
Flashcards
PACER
Exercises
Quizzes
U
0%
M
0%
⚙
Study Tutor
Ask about any concept, get a grounded answer with citations, or say “quiz me” to practise.
＋
New chat
Clear
Explain the agentic loop
Quiz me on Tool Design
What changed with effort vs budget_tokens?
Search for the latest Claude pricing
The study tutor needs JavaScript.
The rest of the site works without it — browse
concepts
,
quizzes
, and
flashcards
directly.