Concepts

Agent Execution Models: Five Ways to Run Architecture Decisions Background Daemon Architecture Cache Economics: The Load-Bearing Infrastructure Configuration and Feature Flags Context Window Management: The Five-Tier Compaction Cascade Cost and Token Economics Error Handling and Recovery Extensibility Model: How Claude Code Grows Five-Layer Platform Architecture Known Bugs and Regressions Memory Design Model Behavioral Engineering Multi-Agent Architecture Multi-Agent Patterns Observability and Cost Permission and Trust Model Security Model Session Lifecycle: From Boot to Dream Supply Chain & Security Incidents (March-April 2026) Telemetry and Observability Testing and Quality Assurance The Community Analysis Phenomenon The God Function and Technical Debt The MCP Ecosystem The Notification and Hooks Pipeline The Source Leak and Its Consequences The Startup Pipeline The Tool Lifecycle Tool System Architecture

Digests

Claude Code Architecture Overview Memory and Agent Systems Security Threat Model

Entities

--resume Parallel Tool Result Loss Bug .js-to-.ts Import Extension Convention .mcp.json Config File /autofix-pr Command /batch Command Worktree Orchestration /bridge-kick Command /cost Command /effort Interactive Slider /init-verifiers Command /insights Command /loop Esc-Cancel Wakeup Semantics /model Mid-Conversation Switch Warning /proactive Command Alias /recap Command /remember Command /resume Picker Default Scoping /rewind Command /team-onboarding Command /thinkback Command Family 1.6% Direct API Call Ratio 10 Steps Per Iteration 10-Section Summary Template 12 Boring Primitives Framework 120K Token-Stop Configuration 14 Cache-Breaking Vectors 1421-Line While Loop 15 Reverse-Engineered Design Intents 16 Bundled Skills 17% Code Comprehension Reduction (AI-Assisted Developers) 20-Day Post-Leak Community Retrospective 24-Hour Internal Alignment Review 24-Hour Pre-Deployment Alignment Review 250,000 Wasted API Calls Bug 27 Hook Events Across 5 Categories 27-Event Hook Pipeline (5 Safety / 22 Lifecycle) 29% False Claims Rate (Code-Path Hallucination Metric) 3,167-Line God Function Metrics 4-Layer Prompt Pipeline 5 Tool-Authorization Hook Events 5-Stage Compression Pipeline 512K vs 64K Line Count Discrepancy 54 Built-in Tools Breakdown (19 Unconditional + 35 Conditional) 64 Undocumented Gated Modules Question 7-Agent Feature Model 8+ MCP Transport Variants 9 Ordered Context Sources 900KB Silent Prompt Rewrite 93% Permission Approval Rate <analysis> Hidden Scratchpad @MODEL_LAUNCH Annotation @include Directive for Memory Files @modelcontextprotocol/sdk ^1.12.1 ACP (Agent Client Protocol) Composition AI Copyright Paradox AI-Authorship Percentage Escalation AI-Written Code at Scale Critique ANR Detection System APIMessage (Stage 2 API Wire Format) Adaptive Thinking Official Documentation Agent Lifecycle Agent SDK Agent Teams Agent Teams Permission Dialog Crash Fix AgentArmor AgentTool Agentic Verifier Hook Execution Type Agnost AI Aisle Security Glasswing Replication Alpine + oven/bun Dockerfile AnalyticsMetadata_I_VERIFIED_THIS_IS_NOT_CODE_OR_FILEPATHS AnalyticsMetadata_I_VERIFIED_THIS_IS_NOT_CODE_OR_FILEPATHS Andrey Kolkov Ant-Only Bridge Dev Overrides Anthropic 132-Engineer Internal Study (Paradox of Supervision) Anthropic Bun Fork Native String Replacement Anthropic DMCA Failure Anthropic IPO Timeline (Q4 2026) Anthropic VDP 'Informative' Disposition Anti-Distillation Defenses AppState Append-Only Auditability Commitment Append-Only Daily Logs Append-Only Durable State (Principle) April 1-12 Release Sprint April 9 Sub-Agent TTL Regression Architectural Convergence Thesis AskUserQuestionTool Async Generator Query Architecture Auditability Over Query Power Principle Auth Backend Selection Matrix Auto Mode Classifier Auto-Compact Auto-Dream Auto-Memory Auto-Memory / Auto-Dream Mutual Exclusion Auto-Memory Operational Controls Auto-Mode Agent Tool Classifier Overflow Bug Auto-Mode Threat Model (4 Categories) AutoBE AutoBE-Claude-Code Coexistence Model AutoDream Service Autocompact Death Loop Avoid-Creating-New-Files Instruction Softening Away Summary Axios RAT (v1.14.1 / v0.30.4) March 31 Supply-Chain Window BRIDGE_MODE Feature Flag Background Plugin Monitors Background Skill Improvement Bash Bypass Wave (8 fixes) Bash Security Pipeline BashTool Bedrock Auth Hotfix Bedrock/Vertex/Foundry 429 Status-URL Fix Billing Sentinel Bug (cch=00000) Bimodal Token Distribution (Pre/Post v2.1.100) Blake Crosley Blake Crosley 84-Hook Orchestration Layer Blake Crosley Cache Fragility Analysis Blame-Shifting Behavior Catalogue Blocking Budget Boris Cherny BoundedUUIDSet Message Dedup Bridge Control Request Protocol Bridge JWT Authentication Bridge Spawn Modes Bridge Stub (src/bridge/stub.ts) Bridge System Bridge Transport v1 vs v2 Brief Mode Brief Mode (/brief) BriefTool Buddy Sprite Build Scanner Codename Paradox Build-Out Prompt Index Bun Bug #28001 Bun DCE Feature-Gate Pattern Bun Runtime Bun-Level Zig HMAC Request Signing Bun-Specific API Stubs CCRClient CLAUDE.md Category Error (Handbook vs Lock) CLAUDE.md Every-Request Injection CLAUDE.md Instruction-Following Ceiling CLAUDE.md Jailbreak Vector CLAUDE.md Lazy Loading CLAUDE.md as User Context Not System Prompt CLAUDE_CODE_AGENT_COST_STEER CLAUDE_CODE_AUTO_COMPACT_WINDOW CLAUDE_CODE_CERT_STORE CLAUDE_CODE_COORDINATOR_MODE CLAUDE_CODE_DEBUG_LOG_LEVEL CLAUDE_CODE_DISABLE_ADAPTIVE_THINKING CLAUDE_CODE_DISABLE_COMMAND_INJECTION_CHECK (removed) CLAUDE_CODE_DISABLE_GIT_INSTRUCTIONS CLAUDE_CODE_ENABLE_AWAY_SUMMARY CLAUDE_CODE_EXECPATH CLAUDE_CODE_FORK_SUBAGENT Env Var CLAUDE_CODE_MAX_CONTEXT_TOKENS CLAUDE_CODE_MCP_ALLOWLIST_ENV CLAUDE_CODE_SANDBOXED CLAUDE_CODE_SCRIPT_CAPS CLAUDE_CODE_SDK_HAS_OAUTH_REFRESH CLAUDE_CODE_SIMULATE_PROXY_USAGE CLAUDE_CODE_SUBPROCESS_ENV_SCRUB CLAUDE_CODE_USE_ANTHROPIC_AWS CLAUDE_CODE_USE_MANTLE CONTEXT_COLLAPSE (Layer 5 Emergency Compaction) CVE-2025-59536 CVE-2026-21852 CVE-2026-2796 Autonomous JS Engine Exploit CVE-2026-31861 CVE-2026-34714 (Vim tabpanel RCE) CVE-2026-35020 CVE-2026-35021 CVE-2026-35022 Cache Creation JSONL Schema Cache Disabled Startup Warning Cache Economics Cache Invalidation Re-Bill Bug (Session Resume) Cache Sticky Latch Cache TTL Scandal (95-Day JSONL Analysis) Cache Write Cost Premium Cache-Aware Tool Registration Cache-Prefix-Change-Suffix Principle CachePointBlock Cached vs Non-Cached Billing Multiple Callback Hook Execution Type Capability Amplification (Value) Capybara v8 False-Claims Rate Regression Capybara v8 False-Claims Regression Cathedral AI Chain Patching via headUuid/anchorUuid/tailUuid Changelog Change-Monitor Pipeline Channels as MCP Servers Chaofan Shou (@Fried_rice) Chrome Native Host Bridge Chrome Native Messaging Host Claude Code Claude Code Feature Flag Catalog Claude Code Guide Subagent Claude Code Review Claude Code Slash-Command Catalog (~85) Claude Code v2.1.104 Claude Code v2.1.105 Claude Code v2.1.113 Claude Code v2.1.114 Claude Flow V3 vs TeammateTool 92% Overlap Claude Mythos Preview Claude Opus 4.7 Claude Self-Observation Passage Claude-in-Chrome MCP Server Claw Code CliMessage (Stage 1 CLI Internal Representation) Cloud Container Runtime (CCR) Code Authorship Anomaly Codified Behavior Over Model Self-Discipline Coding-Is-Solved vs Brittle-Infrastructure Tension Cognee Command Registry (src/commands.ts) Command System Commander.js CLI Parser Compaction Headroom Reservation (~16.5%) Compaction Laundering Path Compaction Pipeline Complete Cache Bug Catalogue (6 Bugs, 4 Layers) Composable Multi-Mechanism Extensibility (Principle) Compounding /insights Loop Concurrent-Read Serial-Write Execution Model Conditional-Statement Fingerprint Config Schemas ConfigTool Consolidation Lock ContentBlock Polymorphism (9 Variants) Context Collapse Read-Time Projection Context as Scarce Resource with Progressive Management (Principle) Context-Aware Tool Restrictions Context-Is-Architecture Thesis Context-Management Decision Contextual Adaptability (Value) Conversations Outgrow a Single Path Principle Conversations Outlive Context Principle Coordinator Mode Coordinator Permission Branch Copy-on-Write Overlay Filesystem Core-10 Tool Set Correction Memory Injection Cost Tracker Cowork-Style Three-Panel Layout Cross-Terminal /effort Cascade CrowdStrike Glasswing Assessment Cryptographic Turn Signing Curia (Static Authorship Profiler) Cursor (Chat-Integrated Category) Custom Shell Parser DMCA Takedown (8,100 repos) Daniel Nakov Dark Code Thesis Dave Morin Dave Shoemaker David Breunig System Prompt Interactive Map Dead-Fork Entry Replay Bug Default Effort Elevation to HIGH Defense in Depth with Layered Mechanisms (Principle) Defense-in-Depth Independence Assumption Failure Deferred Tool Loading Convergence Deferred Tool Schemas (ToolSearch) Denial Tracking System Denial as Routing Signal Denis Stetskov Deny-First with Human Escalation (Principle) Deny-Rules Bypass (50-Subcommand) Diacritics Input Drop Bug Diminishing Returns Detection DreamTask Dual-Representation Message Invariant Dumb Zone Dynamic require() Ant-Only Escape Hatches ENABLE_PROMPT_CACHING_1H ETH Zurich CLAUDE.md Study (Feb 2026) EXTERNAL_PERMISSION_MODES Array Elliot Arledge Emile Michael Employee Post-Edit Verification EnterPlanModeTool / ExitPlanModeTool Entrypoint System Environment Variable Atlas Epitaxy Desktop Multi-Agent Rebuild Execa Command Injection (VULN-01/02/03) Execution-Feedback-Shaped Code Hypothesis Experimental Dreaming System (OpenClaw) Explore Agent Read-Only System Prompt Explore Subagent Extended-Thinking Rotating Indicator Extensibility Decision Externalized Programmable Policy (Principle) FFmpeg H.264 Integer Overflow 16-Year Bug FORCE_PROMPT_CACHING_5M Feature Flag Environment Variable Namespace Feature Flags (Build-Time) Feature-Flag Sprawl Thesis Feature-Gate 60-Minute Hot-Reload Feature-Gated Agent Tool Input Schema Fennec File-History Checkpoints FileEditTool FileReadTool FileWriteTool Fingerprinting AI Coding Agents (MSR '26 Paper) Firefox 148 Exploit Benchmark Five Context-Conservation Mechanisms Beyond Compaction Five-Layer Architecture Five-Layer Compaction Architecture Five-Layer Subsystem Decomposition Fixed Request Assembly Order Focus View (Ctrl+O) Forbes Dark Code Framing Forked Agent Cache-Sharing Economics Forked Agent Pattern Four Memory Systems Four-Phase Cache TTL Timeline Four-Phase Consolidation Prompt FreeBSD NFS RCE CVE-2026-4747 Fresh Context Validator Pattern Fresh-Session Token Baseline (31K) Frustration Telemetry Fucks Chart Function Calling Harness Future Direction: Cross-Session Persistence Substrate Future Direction: Governance and Oversight at Scale Future Direction: Harness Boundary Evolution (Where/When/What/With-Whom) Future Direction: Horizon Scaling (Session to Scientific Program) Future Direction: Long-Term Capability Preservation as First-Class Design Problem Future Direction: Silent Failure and Observability-Evaluation Gap GHSA-mmgp-wc2j-qcv7 (Penligent) GNU Emacs Git Forever-Day Gateway Control Plane Architecture Gather-Act-Verify Loop General-Purpose Subagent Git-Status Cache Bust GitHub Copilot (Inline Completion Category) GitHub Issue #35584 GitHub Issue #38335 GitHub Issue #39755 GitHub Issue #40524 GitHub Issue #42187 (Fix-Loop Token Maximization) GitHub Issue #42796 GitHub Issue #44461 (Systematic Rule Violation Taxonomy) GitHub Issue #45550 (Test-Skip Normalization) GitHub Issue #46917 (Phantom Tokens) GitHub Issue #47098 (Git Cache-Bust) GitHub Issue #528 GitHub Webhook Subscription Glasswing Founding Partners Glasswing Paradox GlobTool Go Rewrite Proposal (Issue #41981) God Function Responsibilities Graceful Recovery and Resilience (Principle) Graceful-Degradation Service Wiring Pattern Graduated Context-Cost Extensibility Ordering Graduated Layering Over Monolithic Mechanisms Graduated Lazy-Degradation Principle Graduated Trust Spectrum (Principle) Grep Over RAG GrepTool GrowthBook GuardContentBlock H100 Teardown Incident HISTORY_SNIP Snip Compaction HTTP Webhook Hook Execution Type Harness Underperformance Gap (77% vs 93%) Harness-Is-The-Differentiator Thesis Harness-as-Differentiator Thesis Harrison Security Deep Dive (Query Loop) Harrison Security query.ts Deep Dive Hooks System Human Decision Authority (Value) Hung Nguyen (Calif AI Red Teaming) HybridTransport (v1 Bridge) IDE/Device Handoff Commands In-App Custom Agent Builder In-Process Isolation Mode InProcessTeammate Ink Render Pipeline Integrated Terminal / Editor / Diff / Preview Panes IntentGuard Interactive Permission Branch Internal Debug Command Cluster Isolated Subagent Boundaries (Principle) Issue Tracker Automation JSONL Token Undercount (100-174x) Jarred Sumner Jeongho Nam Ji_ai 15 Design Intents Analysis Jiacheng Liu Jonny Teardown Judgment-Encoding Hooks Justacuriousengineer Analysis KAIROS KAIROS Exclusive Tools KAIROS Security Concerns KAIROS-AutoDream Contract Kangraemin MITM Analysis (Claude Inspector) Kolkov Watchdog Patch LLM-Based Memory File Selection LLM-Prompt Hook Execution Type LLM-as-Untrusted-Component LSP Service LSP which Fallback Injection LSPTool Language-Hop Transpilation DMCA Evasion Late Context Injection Lazy Skill Invocation Pattern Lazy-Loaded Heavy Dependencies Leak 1: Launch-Day Source Map (Feb 24, 2025) Leak 2: @anthropic-ai/claude-agent-sdk Bundle (Mar 7, 2026) Leak Test-Absence Observation Leaked GitHub Token Scrubbing Procedure Linux Kernel 23-Year Vulnerability Linux PID Namespace Sandbox Local CLAUDE.md LocalAgentTask LocalMainSessionTask Long-Term Capability Preservation (Evaluative Lens) Long-Term Capability Preservation Lens Loop State Record MACRO Global Object MACRO.ISSUES_EXPLAINER MCP Auth MCP Client MCP Ecosystem 17000-Server Milestone MCP Integration MCP Reconnect Memory Leak (v2.1.97 fix) MCP Runtime Management MCP Server Discovery MCP Startup Tax MCP Subsystem MCP Tool Result Post-Hook Ordering MCPTool Magic Docs MagicDocs Service Managed CLAUDE.md Manifest-First Plugin System (OpenClaw) March 2026 Three-Layer Quota Crisis March 23 Rate Acceleration March 6 Cache TTL Regression Melon Mode MemPalace Memory Correction on Rejection Memory Extraction Prompt Memory Hierarchy Memory as Hint, Not Truth Message Normalization Minimal Scaffolding, Maximal Operational Harness (Principle) Missing .npmignore Source Map Rule Model Codenames Model Judgment Within Deterministic Harness Module Size Map Mohit Tripathi Monitor Tool Multi-Repo Support (Epitaxy) Mythos Firefox 147 Exploit Benchmark Mythos Introspective-Curiosity Opening Pattern Mythos Obfuscation Incident Pattern Mythos Self-Covering-Tracks Behavior Nano Claude Code v3 Nate Jones Native Attestation Native Binary Spawn Architecture Native Claude Code Binary Spawn Nicholas Carlini Nine Continue Points Nine-Step Per-Turn Pipeline No-Change-For-Users Post-Leak Finding No-Planner Architecture Decision NodexisLabs Authorship Analysis Non-restoration on Resume Layer NotebookEditTool Numbat Numeric Length Anchors (Anthropic-Only) OAuth Service OS CA Certificate Store Trust OTel Telemetry Spec (April 2026) OWASP ASI (Agentic Security Integrity) On-Demand Language Grammar Loading OpenBSD TCP SACK 27-Year Bug OpenCVE Claude Code Catalogue (23 CVEs) OpenClaw OpenClaw Multi-Agent Routing OpenClaw Shutdown OpenClaw Subscription Ban (April 4, 2026) OpenClaw Workspace Bootstrap File Set Operon Desktop Mode Optional-vs-Essential Dependency Triage Opus 4.7 Context Window Bug (200K vs 1M) Opus 4.7 Tokenizer Ouroboros Architecture Output-Efficiency Internal-vs-External Prompt Gap POSIX flock() Multi-Instance Coordination PROACTIVE Mode Paradox of Supervision Parallel Sessions Sidebar Parallel Startup Prefetch Patterns-Borrowed-Not-Understood Thesis Peers (CC-Codex Session Sharing) Per-Tool-Result Budget Perforce Mode Perforce VCS Support Perimeter-Level Access Control (OpenClaw) Permission Mode Constraints Layer Permission Pipeline Permissions Never Restored on Resume Peter Steinberger Phoenix Security Phoenix Security Kill Chain Picus Security Glasswing Analysis Piebald (Feature Flags) Piebald claude-code-system-prompts Repo Plan Subagent Planning-Execution Session Separation Plugin 4-Component Distribution Bundle Plugin Manifest 10 Component Types Plugin System Podoliako Myth-Busting Analysis Policy Limits Service Post-Leak Claude Code Playbook Post-Leak Default-Prompt Counter-Prompt Pattern PowerShellTool Pre-Prompt Semantic Injection Pattern Pre-Trust Initialization Window Pre-trust Execution Window PreCompact Hook Prevent Sleep (macOS) Principal Hierarchy (Anthropic > Operators > Users) Project CLAUDE.md Project Epitaxy Confirmation Project Glasswing Prompt Cache Section Registry Prompt Priority Hierarchy Proxy Verification Methodology Q1 2026 Feature Release Cluster Query Loop State Object QueryEngine (Conversation Wrapper) QueryEngine 46K-Line File QueryEngine.ts RAPTOR REPL.tsx REPLTool Ralph Wiggum Technique Randal Olson / n_hiraoka Qiita Analysis Rate Limit Traffic Light System Rate Limit Transparency System Rationalization Meta-Pattern ReAct Pattern React Compiler (Enabled) Reactive Compact Circuit Breaker Read-time Projection Non-Destructive Property Read:Edit Ratio Collapse Reasoning-Placement Decision Rebuild Verification Sentinel Pattern Recca Tsai Recurring Design Questions Framework Reduce-Choices-Not-Prohibit Principle Regression Feedback Loop Release Velocity Paradox Reliable Execution (Value) Remote Control Sub-Agent Transcript Streaming Remote Isolation Mode Remote Managed Settings RemoteAgentTask RemoteTriggerTool ReplBridgeHandle Interface Repository Triage Benchmark (5-Tier) Retry System Reversibility-Weighted Risk Assessment (Principle) Round 29 Leak Document Round 30 Post-Leak Culture Analysis Corpus Routines (Headless Claude Code Automations) SDK query() Cleanup Fix SSE Linear Processing Fix SSETransport SYSTEM_PROMPT_DYNAMIC_BOUNDARY Safety, Security, and Privacy (Value) Safety-Posture Decision Sandbox Dangerous-Path Bypass (rm/rmdir) ScheduleCronTool Science Vertical Product Lineage ScopedMcpServerConfig Type Scratchpad (tengu_scratch) Screenshot Base64 Cost Sebastian Raschka Second-Generation Ceiling Self-Contained Worker Context Principle SendMessageTool Server-Side Token Injection (v2.1.100+) Service Layer Session JSONL Log Format Session Memory Service Session Persistence Session Resume Fixes (v2.1.97) Session-Ingress Service Session-Persistence Decision SessionMemory Service Seven Independent Safety Layers Seven-Component High-Level Architecture Shared Failure Modes in Defense-in-Depth Shell Hook Execution Type Side Chats (Command+;) Sigma Detection Pack (16 rules) Sigrid Jin Silent Model Downgrade Six Orchestration Patterns Skill System Skill Tool Built-in Slash Command Invocation SkillTool vs AgentTool Architectural Contrast Slash Command Execution Surface Sleep-Time Compute SleepTool Source Map Leak (March 31, 2026) Source-File-Mapped 7-Component Architecture Southbridge Research Speaker Attribution Bug Spec-Driven Delegation Speculation Engine Speculation Safety Envelope Speculative Classifier Permission Branch Speculative Execution (Full) Stalled Stream Auto-Abort Star Chamber Startup Context Tax Crisis State-Machine Loop Pattern (Not Recursion) Statusbar Context JSON (Accurate Token Source) Statusline-Setup Subagent Straiker Straiker Compaction-Persistence Attack Model StreamAccumulator (Stage 3 Streaming Representation) Streaming Tool Execution 40% Speedup Streaming Tool Execution Parallelism Streaming Watchdog Bug Chain StreamingToolExecutor Strict Write Discipline Sub-Agent Context Isolation Property Sub-Agent TTL Binary Cutover Pattern Sub-Agents as Parallel Design Perspectives Pattern Subagent Permission Override Precedence Subagent Sidechain Transcripts (sessionStorage.ts:247) Subagent-Architecture Decision Subhasish Satapathy Summary-Only Subagent Return Model Supply Chain Cascade Timeline (March 24 - April 3) Swarm Worker Permission Branch SyntheticOutputTool System Prompt Assembly System Prompt Fingerprinting System Prompt Simplicity Bias System Reminders Tab-Acceptance Pipelining Task Master Task System Team Memory Sync Service Team-Lead Polling Loop TeamCreateTool / TeamDeleteTool TeamPCP Supply Chain Campaign TeammateTool TechTrenches God-Function Analysis Telemetry Default-On Asymmetry by Endpoint Telemetry System Tengu Terminal Renderer ThemeProvider Component Thinking Depth Regression (pre-redaction) Thinking Hash System Thinking Hints System Thinking Signature Field Proxy Thinking-Signature Resume Tax Thread Taxonomy Three Extension Injection Points Three Meta-Patterns of Claude Code Three Persistence Channels Three Slash-Command Types Three View Modes (Verbose/Normal/Summary) Three-Block Prompt Cache Layout Three-Generation Agent Taxonomy Three-Layer Claude Code Extension Stack Three-Layer Compression Hierarchy Three-Layer Frustration Detection System Three-Layer Tool Result Flow Three-Layer Verification Three-Model Subagent Execution Matrix Three-Stage Message Pipeline Three-Vector Leak Pattern Tick Prompt System Tiered Enforcement Architecture TodoWriteTool Token Doubling Effect Token Estimation Service TokenCost.app Tokenizer-Effort-Cache Cost-Multiplier Model Tool Documentation Deduplication Tool Interface Contract Tool Pool Deduplication Precedence Tool Pre-filtering Safety Layer Tool Registry Gating Taxonomy Tool Result Budget Tool Result Storage Tool System Tool.ts Tool.ts (Built-in Tool Registry) ToolSearch System Touchpoint Desktop Automation Transparent File-Based Configuration and Memory (Principle) Trusted Device Token Two-Tier Subagent Permission Scoping Two-Version Output Efficiency Directive UDS Inbox (Fabricated Feature) ULTRAPLAN ULTRAPLAN Auto-Environment Creation ULTRAPLAN State Machine (running/needs_input/plan_ready/approved/rejected) ULTRAPLAN Three Variants USER_TYPE === 'ant' Gate UltraPlan Plan-Approval Branching UltraPlan Third Execution Context Ultraplan Web-Fallback Hide Fix Undercover Mode User CLAUDE.md VILA Lab (MBZUAI) VILA Paper Publication Date (April 16, 2026) Values Over Rules (Principle) Velocity-Pressure Code Pattern Verification Agent Gate Verification Subagent Vertex AI Setup Wizard Victor Antos Victor Antos Self-Referential Analysis Vim Mode Virtual Scroller Memory Leak Fix Vitest Test Infrastructure Voice System W3C TRACEPARENT Propagation WebFetch CSS/JS Stripping WebFetchTool WebSearchTool WebSocket 147K/Week Poll-404 Reconnect Willow Mode WorkSecret Payload Worktree Isolation Worktree Isolation Mode Wrong-Edit Retry Loop X42 Protocol Xiaohan Zhao Xinyi Shang YukonSilver VM Yurukusa Zain Hasan Architecture Document Zhiqiang Shen Zig-Based Attestation Client Zilliz acceptEdits Permission Mode additionalDirectories Setting agent-super-spy assemble() Injection Point auto Permission Mode autoCompact.ts awaitAutomatedChecksBeforeDialog Flag bashSecurity.ts Numbered Checks bridge/bridgeMain.ts bubble Permission Mode buddy/ Directory buildTool Factory buildTool Fail-Closed Defaults bun build --compile Standalone Binary bun:bundle Runtime Resolution Problem bun:bundle Runtime Shim bunfig.toml preload + alias Configuration bypassPermissions Permission Mode cache_edits Endpoint cc-relay cc-relay MITM Proxy cc-trace cc:// Protocol ccaudit ccdiag cchistory ccunpacked.dev ccusage claude server / claude connect / claude open Commands claude-code-log claude-code-proxy (seifghazi) claude-pragma claw-code Cargo Workspace (6-Crate) claw-code PARITY.md claw-code usize::MAX Iteration Policy commands.ts commands.ts (Slash Command Implementations) commit Skill Misfire Fix compaction/ Directory constants/system.ts context.ts (Runtime Context Collector) coordinatorMode.ts cost-tracker.ts createMovedToPluginCommand deep-dive-claude-code.vercel.app default Permission Mode disableSkillShellExecution dist/cli.mjs (Production Bundle Artifact) dist/cli.mjs Production Bundle dontAsk Permission Mode esbuild Build System (scripts/build-bundle.ts) execute() Injection Point extractMemories.ts firstPartyEventLoggingExporter.ts forceRemoteSettingsRefresh Policy (v2.1.92) getSimpleDoingTasksSection() handleStopHooks kairos/ Directory keepCodingInstructions Flag llm-interceptor (formerly claude-code-inspector) lord.technology lord.technology Community Audit (April 11) main.tsx main.tsx Entry Point marckrenn/claude-code-changelog mcp-server Sub-Project mcp-server/ Sub-Project model() Injection Point oh-my-codex parity_audit.py pendingToolUseSummary Speculation piebald-ai/cc-prompts plan Permission Mode print.ts process.env.USER_TYPE Build-Time Branch promptSuggestion.ts prompts.ts (141-Module Assembly) query.ts query.ts 10-Step Loop query.ts Seven-Point Continue Map queryLoop() Shared Execution Path recapDetailLevel Config Setting replLauncher.tsx rolloutConfig sandbox.network.deniedDomains Setting scripts/build-bundle.ts (esbuild Production Config) scripts/ci-build.sh scripts/dev.ts Development Launcher scripts/test-auth.ts scripts/test-commands.ts scripts/test-ink.tsx scripts/test-tools.ts services/api/bootstrap.ts services/api/claude.ts (Message API Wrapper) services/api/client.ts showThinkingSummaries Flag src/bridge/ Directory src/components Directory src/context/ Directory (Additional Context Modules) src/coordinator Directory src/entrypoints/ Directory src/entrypoints/cli.tsx src/entrypoints/init.ts (Service Init Sequence) src/entrypoints/mcp.ts (MCP Server Mode) src/ink Custom Fork src/ink Custom Ink Wrapper src/ink Directory src/keybindings Directory src/memdir/ (Memory Directory System) src/migrations Directory src/services/mcp/ Directory src/shims Directory src/tools Directory Layout src/vim Directory startBackgroundHousekeeping tengu_amber_quartz_disabled tengu_api_query messageLength Telemetry tengu_iron_gate_closed tengu_miraculo_the_bard tengu_moth_copse tengu_penguins_off tests/ Directory Layout toolExecution.ts 14-Step Pipeline tsconfig.json bun:bundle Path Mapping tt() Function undercover.ts userPromptKeywords.ts v2.1.100 v2.1.101 v2.1.107 Release v2.1.108 Release v2.1.109 Release v2.1.113 v2.1.114 v2.1.115 v2.1.116 v2.1.117 v2.1.20 Transparency Crisis v2.1.83 Prompt Changes v2.1.89-91 Patch Wave v2.1.98 verificationAgent x402 HTTP Micropayment Mechanism xhigh Effort Level yoloClassifier.ts ~7x Token Cost of Agent Teams

Sources

Advanced setup Agent SDK overview Agent SDK reference - Python Agent SDK reference - TypeScript Agent Skills in the SDK Architecture Authentication Automate work with routines Automate workflows with hooks Best Practices for Claude Code Bits Bytes NN Architecture Analysis Bridge Layer (VS Code / JetBrains IDE Integration) Build-Out Prompt Index CLI reference Changelog Channels reference Checkpointing Choose a permission mode Claude Code Agent Guidance Claude Code Codebase - Deep Technical Understanding Guide Claude Code Docs: architecture Claude Code Docs: commands Claude Code Docs: subsystems Claude Code Docs: tools Claude Code GitHub Actions Claude Code GitLab CI/CD Claude Code Leak Round 30 — Native Binary Shift, Desktop Multi-Agent, Opus 4.7 Tokenizer Economics (2026-04-21) Claude Code Leak Round 31 — VILA Paper, Proxy Token Injection, v2.1.116/117, Three-Stage Pipeline (2026-04-21 to 2026-04-23) Claude Code README (razakiau fork) Claude Code Skill Definition Claude Code in Slack Claude Code on Amazon Bedrock Claude Code on Google Vertex AI Claude Code on Microsoft Foundry Claude Code overview Claude Code settings Claude Code with GitHub Enterprise Server Code Review Commands Commands Reference Common workflows Community Deep Dive: Architecture, KAIROS, Auto-Dream, Agent Teams Configure permissions Configure server-managed settings Connect Claude Code to tools via MCP Connect to external tools with MCP Constrain plugin dependency versions Continue local sessions from any device with Remote Control Create and distribute a plugin marketplace Create custom subagents Create plugins Customize keyboard shortcuts Customize your status line Data usage Deep Dive Claude Code DeepWiki Claude Code Analysis Development containers Discover and install prebuilt plugins through marketplaces Dive into Claude Code — Architecture Deep Dive Dive into Claude Code — Design Guide for Agent Builders Dive into Claude Code — Related Resources Dive into Claude Code — Repo Overview Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems (arXiv:2604.14228) Enterprise deployment overview Enterprise network configuration Environment variables Error reference Exploration Guide Explore the .claude directory Explore the context window Extend Claude Code Extend Claude with skills Find bugs with ultrareview Fullscreen rendering Get started with Claude Code on the web Get started with the desktop app Get structured output from agents Give Claude custom tools Handle approvals and user input Hooks reference Hosting the Agent SDK How Claude Code works How Claude remembers your project How the agent loop works Interactive mode Intercept and control agent behavior with hooks Internals: Auto-Memory, Auto-Dream, and Agent Teams JetBrains IDEs LLM gateway configuration Legal and compliance Let Claude use your computer from the CLI Manage costs effectively Migrate to Claude Agent SDK Model configuration Modifying system prompts Monitoring Observability with OpenTelemetry Optimize your terminal setup Orchestrate teams of Claude Code sessions Output styles Plan in the cloud with ultraplan Platforms and integrations Plugins in the SDK Plugins reference Prompt 01: Install Bun Runtime & Dependencies Prompt 02: Runtime Shims for `bun:bundle` Feature Flags & `MACRO` Globals Prompt 03: Create esbuild-Based Build System Prompt 04: Fix MCP Server Build Prompt 05: Environment Configuration & API Authentication Prompt 06: Verify and Fix the Ink/React Terminal UI Pipeline Prompt 07: Audit and Wire Up the Tool System Prompt 08: Audit and Wire Up the Command System Prompt 09: Get the QueryEngine (Core LLM Loop) Functional Prompt 10: Wire Up System Prompt, Context Gathering & Memory System Prompt 11: MCP Client/Server Integration Prompt 12: Wire Up Services Layer (Analytics, Policy, Settings, Sessions) Prompt 13: Bridge Layer (VS Code / JetBrains IDE Integration) Prompt 14: Create Development Runner Prompt 15: Production Bundle & Packaging Prompt 16: Add Test Infrastructure & Smoke Tests Push events into a running session with channels Quickstart Rewind file changes with checkpointing Round 10: Quality Gap, CVE, Grep Over RAG Round 11 The Fabricated Features, the @MODEL_LAUNCH Pattern, Phoenix Security's Round 12 Background Daemon Architecture, Cache Landmines, Deny-Rule Bypass, and Round 13 The tt() God Function, Frustration Telemetry, X42 Payments, the DMCA F Round 14 REPL.tsx's 227 Hooks, Piebald's 141-Version System Prompt Evolution, T Round 15 The Bun Bug Root Cause, query.ts Inner Loop, the Copyright Paradox, Co Round 2 Community Intelligence Roundup Round 25 - Deep Codebase Analysis (April 9-11, 2026) Round 26 - Speculation Engine Deep Dive (April 11-12, 2026) Round 27 - Speaker Attribution & Thinking Depth (April 12-13, 2026) Round 28 - 1,421-Line While Loop & Env Atlas (April 13, 2026) Round 29 - Cache TTL, Epitaxy Confirmed, Sub-Agent Patterns (April 15, 2026) Round 3 Code Quality, Prompt Architecture, BUDDY, Mythos & the AI Whistleblower Round 3 — New Community Findings, April 1 2026 Round 4 Architecture Deep Dives & Community Synthesis Round 5 Evening Discoveries, April 1 2026 Round 7: The Deepest Architecture Yet Round 8 Design Philosophy, Community Discoveries, and the Engineering Behind th Round 9 What the Community Found Between April 1 and April 3 (1) Run Claude Code programmatically Run prompts on a schedule Sandboxing Scale to many tools with tool search Schedule recurring tasks in Claude Code Desktop Securely deploying AI agents Security Slash Commands in the SDK Speed up responses with fast mode Stream responses in real-time Streaming Input Subagents in the SDK Subsystems Guide The Architecture and Engineering of Claude Code Internals Todo Lists Tools Reference Track cost and usage Track team usage with analytics Troubleshooting TypeScript SDK V2 interface (preview) Use Claude Code Desktop Use Claude Code features in the SDK Use Claude Code in VS Code Use Claude Code on the web Use Claude Code with Chrome (beta) V1 Wiki Index (migration source) V1 Wiki: agent-lifecycle V1 Wiki: agent-loop V1 Wiki: anti-distillation V1 Wiki: api-client V1 Wiki: app-state V1 Wiki: auto-dream V1 Wiki: auto-memory V1 Wiki: bash-security V1 Wiki: bridge V1 Wiki: cache-economics V1 Wiki: channels-daemon V1 Wiki: claude-md V1 Wiki: configuration V1 Wiki: context-compression V1 Wiki: context-engineering V1 Wiki: convergence V1 Wiki: engineering-culture V1 Wiki: error-handling V1 Wiki: five-layer-architecture V1 Wiki: fleet-economics V1 Wiki: forked-agent-pattern V1 Wiki: grep-over-rag V1 Wiki: hooks-system V1 Wiki: kairos V1 Wiki: known-bugs V1 Wiki: mcp-integration V1 Wiki: memory-architecture V1 Wiki: memory-hierarchy V1 Wiki: message-normalization V1 Wiki: multi-agent V1 Wiki: permission-pipeline V1 Wiki: robots-first-engineering V1 Wiki: scheduled-tasks V1 Wiki: session-persistence V1 Wiki: speculation-engine V1 Wiki: system-prompt-assembly V1 Wiki: terminal-rendering V1 Wiki: the-leak V1 Wiki: token-budget V1 Wiki: tool-system Voice dictation Week 13 · March 23–27, 2026 Week 14 · March 30 – April 3, 2026 Week 15 · April 6–10, 2026 What's new Work with sessions Zero data retention Deep Codebase Analysis & Community Feedback Round 22 (April 8, 2026) Deep Codebase Insights & Community Feedback Round 24 (April 9, 2026) Round 16 New Community Insights (April 5–6, 2026) Round 17 April 6, 2026 Round 18 April 6, 2026 (Evening) Round 19 April 7, 2026 Round 19 Late April 6, 2026 Round 20 April 7–8, 2026 Round 20 April 8, 2026 Round 21 April 8, 2026 (Morning) Round 23 Community Feedback, Quality Fixes & Codebase Insights (April 9, 2026

Home › Entities › WebFetch CSS/JS Stripping

WebFetch CSS/JS Stripping

Entity ID: ent-20260419-3d81f23b0f1b
Type: concept
Scope: shared
Status: active
Aliases: webfetch preprocessing

Description

v2.1.105 preprocessing step: strips