Category Index¶

Articles grouped by category and sub-topic.

Concepts¶

concepts/ai-for-small-business — Frameworks and patterns for integrating LLMs into small business/contracting
concepts/claude-code — Anthropic's AI-assisted development tool; enables non-programmers and professionals to build with agents; includes Claude Cowork and Office integrations
concepts/claude-code-skills — Modular SKILL.md-based capabilities that extend Claude Code; bundled skills (/batch, /loop, /simplify), frontmatter reference, invocation control, and the agentskills.io open standard
concepts/context-files — The CLAUDE.md / AGENTS.md / DESIGN.md / SKILL.md family of version-controlled agent context files; persistent project knowledge for AI agents

concepts/agent-harness — The complete infrastructure wrapping an LLM: 12 components (orchestration, tools, memory, context, state, errors, guardrails, verification, subagents), framework implementations, scaffolding metaphor, 7 design decisions
concepts/agentic-engineering — Professional AI-assisted coding: TDD patterns, templates, hoarding techniques, and managing parallel agents
concepts/agentic-workflows — Anthropic's five composable LLM workflow patterns and when to use true autonomous agents
concepts/ai-inflection-point — November 2025 reliability threshold; dark factory pattern; future of software engineering roles
concepts/harness-engineering — OpenAI Codex team's agent-first engineering model: zero manually-written code, 1M lines, context management, enforced architecture, entropy cleanup
concepts/multi-agent-misalignment — Emergent failure where individually aligned agents collectively produce false institutional records; role-fidelity causes narrative drift; state keeper agent as proposed fix

concepts/llm-knowledge-base — Karpathy's LLM-compiled wiki pattern: raw/ → wiki/ → Q&A → outputs; index-based, no RAG needed at moderate scale
concepts/obsidian-claude-code-os — Obsidian + Claude Code as personal AI OS: vault-as-context, relationship graph access via Obsidian CLI, custom slash commands, inline delegation
concepts/personal-knowledge-management — Atomic notes, knowledge graphs, AI skills; pipeline from analog reading to active tools

concepts/ai-red-teaming — Microsoft AIRT's 8 lessons from 100+ product red-teaming operations; threat model ontology, XPIA, PyRIT, RAI harms; Mythos exploit benchmarks
concepts/frontier-ai-cyber-capabilities — AISI evaluation of 7 frontier models on attack scenarios; 6× capability improvement in 18 months; £65/attack; NCSC defender advantage and baseline guidance
concepts/llm-tier-security — Personal security posture for the Mythos era: network isolation, hardware keys, Linux sandboxing, outbound firewalls, financial account alerting
concepts/prompt-injection — Unsolved LLM security vulnerability; lethal trifecta; XPIA; Crescendo; Challenger disaster prediction
concepts/openclaw-security — CVE-2026-33579 privilege escalation; OpenClaw's inherent lethal-trifecta risk; assume-compromise guidance

concepts/mcp-authentication — OAuth 2.1 MCP auth with Entra ID, FastMCP RemoteAuthProvider, VS Code pre-registration, and OBO flow for downstream APIs

entities/andrej-karpathy — Coined "vibe coding"; described LLM knowledge base pattern; authored canonical LLM Wiki gist (Apr 2026)
entities/chuck-kyle — Contractor-turned-digital-marketer and AI early adopter
entities/ethan-mollick — Wharton professor, author of One Useful Thing; Models/Apps/Harnesses framework
entities/google-stitch — Google team behind the DESIGN.md open specification for agent-readable design rules
entities/harmeet-dhillon — DOJ Assistant Attorney General; announced DOJ challenge to Colorado SB24-205 AI regulation (2026-04-25)
entities/pamela-fox — Python/Azure developer; author of guide on authenticated MCP servers with Entra ID
entities/rohit-krishnan — Researcher/writer; demonstrated multi-agent narrative drift via Helios Field Services / Vei experiment (2026-04-24)
entities/simon-willison — Django co-creator, coined prompt injection, leading agentic engineering practitioner
entities/vannevar-bush — Inventor of the Memex (1945); historical antecedent to LLM knowledge bases
entities/vei — Virtual enterprise simulation framework by Strange Lab; substrate for multi-agent misalignment research

concepts/ai-regulation — U.S. AI regulatory landscape; DOJ vs. Colorado SB24-205; federal preemption of state AI law

guides/ai-tool-selection — Framework for choosing AI tools in the agentic era: Models/Apps/Harnesses, leading models, tool recommendations by use case and level

guides/local-agent-stack — Five-layer stack for fully local autonomous agents: llama.cpp, quantization, llama-server, ChromaDB, LangGraph

guides/openclaw-docker — Running OpenClaw in Docker safely with Telegram integration and web UI

guides/progressive-web-scraping — Four-tier escalating scraper: WebFetch → cURL → Playwright → Bright Data MCP; free for 95% of sites