Wiki Log¶
Append-only chronological record of all wiki operations. Each entry uses the format:
Operations: ingest, query, lint, update, init, output
Parse recent entries: grep "^## \[" wiki/log.md | tail -10
[2026-04-03] init | Wiki initialized¶
Created directory structure, empty index files, and meta/changelog.md.
[2026-04-03] ingest | Chuck Kyle YouTube transcript¶
Created concepts/ai-for-small-business, concepts/claude-code, entities/chuck-kyle; updated all indexes and glossary.
[2026-04-03] update | Chuck Kyle YouTube transcript (re-processed)¶
Expanded ai-for-small-business with internet analogy, "ask Claude first" principle, contractor photo advice, bookkeeping example, dashboard model prediction, voice-to-text section; updated glossary (added 4 terms); updated chuck-kyle entity with speech-to-text tool URL.
[2026-04-03] ingest | Simon Willison / Lenny's Podcast transcript¶
Created concepts/agentic-engineering, concepts/ai-inflection-point, concepts/prompt-injection, entities/simon-willison; updated concepts/claude-code with web/YOLO mode details; updated all indexes, glossary (added 10 terms), open questions (added 9 questions).
[2026-04-03] ingest | OpenAI harness engineering article + Simon Willison openclaw-docker TIL¶
Created concepts/harness-engineering, guides/openclaw-docker; updated concepts/ai-inflection-point and concepts/prompt-injection with cross-references; resolved 1 open question; updated all indexes, glossary (added 5 terms), open questions (added 4 questions).
[2026-04-03] ingest | dSebastien analog PKM article + Karpathy LLM knowledge base thread¶
Created concepts/llm-knowledge-base, concepts/personal-knowledge-management, entities/andrej-karpathy; updated all indexes, glossary (added 7 terms), open questions (added 7 questions).
[2026-04-03] ingest | SitePoint local agent stack tutorial + pvergadia Vibe Code Framework tweet¶
Created guides/local-agent-stack; added product-oriented vibe coding section to concepts/agentic-engineering; updated all indexes, glossary (added 3 terms), open questions (added 3 questions).
[2026-04-03] ingest | Greg Isenberg / Internet Vin YouTube transcript + @gregisenberg tweet thread¶
Created concepts/obsidian-claude-code-os; updated concepts/personal-knowledge-management cross-reference; updated all indexes, glossary (added 3 terms: Inline Delegation, Obsidian CLI, Reference Files), open questions (added 4 questions).
[2026-04-03] ingest | Dan Goodin / Ars Technica OpenClaw CVE-2026-33579 + @pvergadia Vibe Code visual¶
Created concepts/openclaw-security; updated guides/openclaw-docker with CVE warning, concepts/prompt-injection with CVE reference, concepts/agentic-engineering with Vibe Code visual; updated all indexes, glossary (added CVE-2026-33579 and updated OpenClaw entries), open questions (added 2 questions).
[2026-04-04] ingest | Pamela Fox — Building MCP servers with Entra ID¶
Created concepts/mcp-authentication and entities/pamela-fox; updated all indexes, glossary (added 4 terms: MCP, OAuth 2.1, OBO flow, MI-as-FIC, Pre-registration), open questions (added 5 questions).
[2026-04-04] output | Marp slide deck — AI for Small Business Getting Started¶
Generated outputs/slides/ai-for-small-business-getting-started.md — 15-slide intro to AI for small business owners.
[2026-04-07] ingest | Ethan Mollick — A Guide to Which AI to Use in the Agentic Era¶
Created guides/ai-tool-selection and entities/ethan-mollick; updated concepts/claude-code with Claude Cowork and Office plugin details; updated all indexes, glossary (added 3 terms: Claude Cowork, Models/Apps/Harnesses, NotebookLM), open questions (added 4 questions).
[2026-04-07] ingest | Claude Agent Skills Explained (YouTube fOxC44g8vig — first pass)¶
Created concepts/claude-code-skills; updated concepts/claude-code with skills section; updated all indexes, glossary (added 7 terms), open questions (added 4 questions, resolved 1).
[2026-04-07] update | Claude Agent Skills Explained (re-ingested with actual transcript)¶
Enriched concepts/claude-code-skills with progressive disclosure model (30–50 tokens/skill), portability claim, four-layer architecture table, multi-subagent skill sharing; resolved 2 open questions.
[2026-04-07] output | Marp slide deck — Claude Code Skills for Small Business Startups¶
Generated outputs/slides/ — slide deck on Claude Code skills aimed at small business audience.
[2026-04-08] ingest | LLM Wiki (Karpathy GitHub Gist, Apr 4 2026)¶
Saved to raw/articles/llm-wiki-karpathy.md. Upgraded concepts/llm-knowledge-base to complete status with canonical source; added entities/vannevar-bush (Memex); updated entities/andrej-karpathy and concepts/obsidian-claude-code-os (Dataview, Web Clipper, image tips); added 7 glossary terms; 1 new open question (qmd); built tools/search.py two-stage search.
[2026-04-08] update | Schema improvements — log format, YAML frontmatter, Q&A filing, ingest discussion step¶
Converted changelog to grep-parseable log.md; updated CLAUDE.md article template to YAML frontmatter; made Q&A filing default behavior; added discussion step to ingest pipeline.
[2026-04-09] ingest | Anthropic building-effective-agents + Microsoft AIRT red-teaming paper + Miessler progressive web scraping¶
Created concepts/agentic-workflows (5 LLM patterns + ACI), concepts/ai-red-teaming (8 lessons, XPIA, PyRIT, Crescendo), guides/progressive-web-scraping (4-tier fallback system); updated concepts/prompt-injection with XPIA + Crescendo + cost-of-attack sections; updated all indexes, glossary (13 new terms), open questions (10 new questions).
[2026-04-15] ingest | NCSC — Why Cyber Defenders Need to Be Ready for Frontier AI¶
Created concepts/frontier-ai-cyber-capabilities with AISI evaluation findings (7 models, 32-step enterprise attack, Claude Opus 4.6 completing 15.6 steps at £65/attempt, 6× improvement in 18 months), defender advantage framework, and AI defensive opportunities; updated concepts/ai-red-teaming cross-reference; updated all indexes, glossary (6 new terms: AISI, NCSC, Shape the Battlefield, Dual-Use AI Capabilities, AIxCC), open questions (6 new questions).
[2026-04-16] ingest | The Anatomy of an Agent Harness (@akshay_pachaar)¶
Created concepts/agent-harness (12 harness components, TAO loop, framework implementations for Anthropic/OpenAI/LangGraph/CrewAI/AutoGen, scaffolding metaphor, 7 design decisions); updated concepts/harness-engineering and concepts/agentic-workflows with cross-links; 5 glossary terms (Agent Harness, Ralph Loop, ReAct Loop, Scaffolding, TerminalBench); 4 new open questions; resolved 1 (Mollick Harnesses = agent harness).
[2026-04-21] ingest | LLM-tier personal computer security (cata, LessWrong)¶
Created concepts/llm-tier-security (threat model shift, Mythos exploit benchmarks, defensive posture across network/OS/financial layers); updated concepts/ai-red-teaming with Mythos tier-5 data; 6 new glossary terms (Mythos, ACATS, Supply-Chain Attack, FIDO2/WebAuthn, OpenSnitch); 5 new open questions; added to Security section of _category_index.
[2026-04-25] ingest | Sub-Agents vs Agent Teams (Suryansh Tiwari, X Article)¶
Updated concepts/agentic-workflows with sub-agents vs. agent teams section, context-based decomposition, Claude Agent SDK example; 3 glossary terms (Sub-Agent, Agent Team, Context-Based Decomposition); 3 open questions; _sources.md, _index.md updated.
[2026-04-25] ingest | Colorado SB24-205, Rohit Krishnan multi-agent misalignment, Google Stitch DESIGN.md, Claude Code Best Practices¶
Added 4 sources: ai-regulation, multi-agent-misalignment, context-files concepts; harmeet-dhillon, rohit-krishnan, vei, google-stitch entities; updated claude-code.md with official Anthropic best practices section; 12 glossary terms; 22 open questions across 4 new sections; all indexes updated.