| Llm Wiki Pattern |
Pattern for LLM-maintained persistent wiki instead of stateless RAG |
5 |
2026-04-24 |
| Rag Vs Llm Wiki |
Comparison: stateless retrieval vs stateful compounding knowledge |
4 |
2026-04-24 |
| Knowledge Compounding |
Principle that each source enriches existing pages, not just adds documents |
3 |
2026-04-24 |
| Three Layer Architecture |
Raw Sources (immutable), Wiki (LLM-maintained), Schema (governance) |
3 |
2026-04-24 |
| Ingest Pipeline |
5-step pipeline: Resolve → Route → Synthesize → Embed → Update |
3 |
2026-04-24 |
| Query Pipeline |
RAG over compiled wiki pages: Embed → Search → Assemble → Stream |
2 |
2026-04-24 |
| Lint Operation |
Health checks: orphans, broken links, contradictions, stale content |
3 |
2026-04-24 |
| Schema Pagespec |
JSON-defined page universe; contract between human intent and LLM execution |
1 |
2026-04-24 |
| Query Templates |
Pre-built prompts for synthesis, gap-finding, debate, output, health, personal |
1 |
2026-04-24 |
| Claude Code Hooks |
Automated triggers (session start, pre-compact, session end) for capturing conversation knowledge |
1 |
2026-04-24 |
| Hot Cache |
~500-char cache of most recent context for quick agent access |
2 |
2026-04-24 |
| Compiler Analogy |
Knowledge processing maps to software compilation: raw→compile→executable→test→runtime |
2 |
2026-04-24 |
| Index And Log |
index.md (content catalog) and log.md (chronological record) for wiki navigation |
2 |
2026-04-24 |
| Vibe Coding |
Natural-language-driven coding where AI generates code from prompts; two modes: pure vs responsible |
5 |
2026-04-24 |
| Ai Slop |
AI-generated code/content that is bloated, buggy, insecure, superficially functional |
2 |
2026-04-24 |
| Agentic Workflows |
Natural language describes outcome, AI agent handles implementation; Wave 3 of AI automation |
1 |
2026-04-24 |
| Code Bloat |
AI tendency to generate excessive code volume; major complexity driver |
2 |
2026-04-24 |
| Cognitive Debt |
Gradual erosion of developer understanding from over-reliance on AI code generation |
2 |
2026-04-24 |
| Security Debt |
Accumulation of vulnerabilities in AI-generated code shipped without proper review |
2 |
2026-04-24 |
| Vibe Deploying |
One-click deployment of AI-generated apps, removing DevOps bottleneck |
1 |
2026-04-24 |
| Hallucination |
LLMs confidently producing incorrect info because they predict tokens, not facts |
2 |
2026-04-24 |
| Overconfidence Effect |
AI-assisted developers produce less secure code but believe it's more secure |
2 |
2026-04-24 |
| Lost Generation |
Risk of engineers lacking deep skills because AI replaces junior-level learning work |
1 |
2026-04-24 |
| Runaway Loops |
AI agents spiraling off-rails with vague instructions, taking catastrophic actions |
2 |
2026-04-24 |
| Production Wall |
Real-world obstacles non-coders face when shipping vibe-coded apps beyond prototype |
1 |
2026-04-24 |
| Three Waves Of Ai Automation |
Evolution: chatbots → AI+automation platforms → agentic workflows |
1 |
2026-04-24 |
| Cat Tools |
Software that helps translators with TM, fuzzy matching, glossaries, file format support |
1 |
2026-04-24 |
| Omegat Architecture |
Component-based, event-driven Java/Swing architecture with modular Gradle build |
1 |
2026-04-24 |
| Plugin System |
OmegaT's .jar-based extensibility for filters, MT engines, glossaries, markers |
1 |
2026-04-24 |
| Filter System |
Two-fold read/write components for parsing many file formats in CAT tools |
1 |
2026-04-24 |
| Translation Memory |
Database of previous translations with fuzzy matching, TMX format, SRX segmentation |
1 |
2026-04-24 |
| Localization Tools Comparison |
Comparison of 6 open-source localization tools across architecture, features, and use cases |
5 |
2026-04-24 |
| Continuous Localization |
CI/CD-integrated translation workflow: automated string extraction, translation, and generation |
3 |
2026-04-24 |
| In Context Editing |
Click text in running app to translate in-place; SDK-based overlay editing |
1 |
2026-04-24 |
| Git First Localization |
Translations live in Git repo as code; repo is single source of truth |
1 |
2026-04-24 |
| Machine Translation |
Automated neural translation; LibreTranslate/Argos as leading open-source option |
1 |
2026-04-24 |
| Pi Coding Agent |
Minimal open-source coding agent by Block: 4 tools, <1000 token prompt, TypeScript extensions |
2 |
2026-04-24 |
| Minimal Agent Harness |
Design philosophy: smallest tool set, trust model's RL training, save context for work |
2 |
2026-04-24 |
| Agent Slop |
Accumulation of low-quality AI-generated code from high-velocity agents without human friction |
2 |
2026-04-24 |
| Agent Skills |
Organized folders of procedural knowledge that agents dynamically load; progressive disclosure |
2 |
2026-04-24 |
| Bayesian Reasoning |
Framework for updating beliefs based on evidence; prior × likelihood / total evidence |
2 |
2026-04-24 |
| Tit For Tat |
Simplest winning strategy for iterated prisoner's dilemma: cooperate first, then copy opponent |
1 |
2026-04-24 |
| Prior Probability |
Probability of hypothesis before evidence; the base rate that evidence updates |
2 |
2026-04-24 |
| Posterior Probability |
Probability of hypothesis after evidence; the updated belief from Bayesian reasoning |
2 |
2026-04-24 |
| Mcp Server |
Open standard for connecting AI agents to external tools and data; "the hands" of agents |
4 |
2026-04-24 |
| Coding Agent Comparison |
Comparison matrix: Claude Code, Pi, Open Code, AMP, Factory Droid |
2 |
2026-04-24 |
| Agent Extensions |
TypeScript-based customization of coding agent behavior, UI, tools, workflows |
2 |
2026-04-24 |
| Progressive Disclosure |
Context management: show metadata first, reveal details only when needed |
2 |
2026-04-24 |
| Argos Translate |
Open-source neural MT library powering LibreTranslate; based on OpenNMT |
2 |
2026-04-24 |
| Responsible Ai Coding |
Using AI code generation while maintaining human oversight and ownership |
2 |
2026-04-24 |
| Ai Assisted Development |
Spectrum from code completion to full autonomous agents |
1 |
2026-04-24 |
| Xliff |
XML Localization Interchange File Format; standard for exchanging translations |
3 |
2026-04-24 |
| Base Rate Fallacy |
Ignoring prior probability when evaluating likelihood; common cognitive bias |
2 |
2026-04-24 |
| Po Files |
Portable Object files; standard format for gettext translations |
1 |
2026-04-24 |
| Gettext |
GNU internationalization framework; uses PO files for translations |
1 |
2026-04-24 |
| Self Hosted Mt |
Machine translation on own infrastructure; privacy, offline, zero per-char costs |
2 |
2026-04-24 |
| Boooos |
Compound errors from AI agents; coined by Mario Zechner |
1 |
2026-04-24 |
| Prisoners Dilemma |
Game where cooperation is better for both but each has incentive to defect |
1 |
2026-04-24 |
| Iterated Games |
Repeated interactions where past behavior influences future decisions |
1 |
2026-04-24 |
| Cooperation Strategy |
Nice + retaliatory + forgiving + clear wins repeated games |
1 |
2026-04-24 |
| Game Theory |
Mathematical study of strategic decision-making among rational actors |
1 |
2026-04-24 |
| Likelihood |
P(Evidence |
Hypothesis); how well hypothesis predicts the evidence |
2 |
| Belief Updating |
Iterative process of revising probabilities as new evidence arrives |
1 |
2026-04-24 |
| Representative Sampling |
Thinking in concrete counts ("4 out of 100") instead of percentages |
2 |
2026-04-24 |
| Procedural Knowledge |
Knowledge of how to do things; packaged in Agent Skills |
2 |
2026-04-24 |
| Vibe Coding Costs |
Hidden and ongoing expenses of AI-assisted development |
1 |
2026-04-24 |
| Authentication Challenge |
Dual-auth systems can't be solved out-of-box by vibe coding tools |
1 |
2026-04-24 |
| Multi Agent Delegation |
Main agent delegates sub-tasks to multiple sub-agents running simultaneously |
1 |
2026-04-24 |
| Geo Optimization |
Optimizing content for AI search engines (ChatGPT, Perplexity, Gemini) |
1 |
2026-04-24 |