Choose where to go

Read the ideas. Build the systems.

Follow a guided Start Here path, read the complete Blog feed, or browse the Learn with Kody catalog for exact prompts, source, and live examples.

Blog stream

Blog

The complete chronological feed stays together here, newest first.

September 7, 2026 1 article
1. Text Me Once: Notification-Agent Design Rules I Learned from a Nine-Day Spam Storm
August 29, 2026 1 article
1. The Interface Should Meet You Halfway
August 19, 2026 4 articles
1. The Trail You Leave Is the Story
2. The Colony Calls the Twin
3. Scrape Yourself
4. The Negative
August 1, 2026 6 articles
1. The Wire Is the Authority
2. Six Times the Test Was Wrong
3. Rarity Without a Ledger: What Adopt Me Actually Taught Me
4. Make the Global Layer Read-Only
5. A Different Answer Is a Measurement
6. A 403 Is an Answer, Silence Is Not
July 14, 2026 1 article
1. The Week Before the Model Leaves
July 5, 2026 2 articles
1. What Is Our Moat? (The Honest Version)
2. The AI You Keep
June 19, 2026 1 article
1. Funded R&D: A 90-Day Plan to Own Your Expertise
June 18, 2026 1 article
1. A Taste of AGI
May 17, 2026 2 articles
1. The leviathan refused to confabulate
2. Cells all the way down
May 6, 2026 2 articles
1. The frozen kernel
2. Public front doors with private brains
May 3, 2026 3 articles
1. Twenty-four words
2. Legacy, Not Delete: Why AI-Generated Systems Need Different Memory
3. Deterministic AI: SHA-256 as a Random Number Generator
May 2, 2026 3 articles
1. The Vibe-Coding-Demo Loop — Parallel Agents, Live Demos, Shipped
2. The mitosis rule
3. Ship the Shape, Keep the Content: The Twin Engine Pattern
May 1, 2026 1 article
1. The founding-100 paradox
April 30, 2026 2 articles
1. Protocol at Ingress, Judgment at Promotion
2. Obsidian as a living twin
April 29, 2026 1 article
1. Autonomy Budget: $200/day, Circuit Breakers, and the Virtue of Hard Caps
April 28, 2026 3 articles
1. Federation via schema adaptation
2. Public Blog Posts Are Training Data: Saying the Quiet Part Out Loud
3. If You Don't Own the File Format, You Don't Own the AI
April 27, 2026 2 articles
1. Don't Let AI Agents Run Python: The Case for a Tiny Lisp Sandbox
2. Daemons in the browser
April 26, 2026 4 articles
1. The Repo IS the Platform
2. The Honeypot Principle: What Your Platform Looks Like When No One Is Watching
3. CSS filter on a 3D wrap silently flattens backface-visibility
4. Endpoints without buttons: when the backend keeps a feature the UI forgot
April 25, 2026 2 articles
1. The Factory Pattern: Why Your AI Factory and Its Outputs Live in Different Repos
2. Three rules I wrote because of three failures
April 24, 2026 11 articles
1. Your Pitch Slide Is a Partner List Misread
2. Writing for the AI Reader — CLAUDE.md, AGENTS.md, skill.md, and the Contract with Assistants
3. Building a Static Site in Pure HTML/CSS/JS in 2026 — The Anti-Framework Case
4. Sacred Files: Declaring Code Frozen on Purpose
5. Roots Are Public Surfaces
6. One URL Per Audience: the Presenter-Mode HTML Slide
7. Markdown Is the Spec; HTML Is the Rendering
8. Three bash scripts run the universe
9. The Directory-README Pattern — Per-Folder Scale Rules as the Rib of Organizational Discipline
10. Let the community govern; don't filter the content
11. The Audience Site Shell — No Build Step Required
April 23, 2026 1 article
1. Two file extensions, one plugin contract
April 22, 2026 3 articles
1. Write Twice: Why Every Public Post Should Have a Private Twin
2. Two Engines, One Hash
3. Seeds as literal seeds
April 20, 2026 1 article
1. When Humans, Agents, and Bots All Push to Main: Rules for Sharing a Repo
April 19, 2026 21 articles
1. Why 5 Templates and Not 50
2. What a Curator does
3. The vendored-repo install pattern
4. The URL Fragment Is the Session
5. The Ten-Prompt Game
6. Pricing for legacy, not utility
7. Permian Reset: Design Doc
8. Patent-pending in build-in-public: how we say what we have without giving it away
9. Open the bones, close the body, sell the soul
10. Offline-resilient send queue
11. The Loop Is the Engine: Why Frame-Based AI Beats Request-Response
12. The integrated-combination doctrine for AI-system patents
13. Heuristics Fail Loudly; LLMs Fail Quietly
14. Git Worktrees Are Not for Experts
15. Five Tones in Sixty Lines
16. The First Currency Problem
17. Filing a software patent in 90 minutes for $160
18. The Engine Hub Pattern — Deploy, Prove, Extract, Park, Reuse
19. The endowment-funded perpetual service
20. Determinism compounds — the math that kills multi-hop pipelines
21. Blog Post as Building Material
April 18, 2026 25 articles
1. Worktrees as Apartments: The Good Neighbor Protocol
2. Trading-Card Visual Treatment as Universal Honeypot
3. Static Sites Are the Right Default for Personal Infrastructure
4. Service-Account Swarms Are Not Impostors
5. The Seed Is the URL
6. Recovery Beats Prevention: The safe_commit Pattern
7. Recovery Beats Prevention, Again: The Binder Permalink Incident
8. Pushing Through the PR Rate Limit
9. Permalinks Are a Promise
10. Own Everything as a Twin
11. Land the Plane: Work Isn't Done Until It's Pushed
12. Inbox Deltas as the Integration Contract
13. Iframe Twins for Surface Compatibility
14. GitHub Issues as a Universal Write API
15. Field Notes Are the Actual Moat
16. Field Notes as a First-Class Artifact
17. Why Agents Don't Use Your Fenced Code Blocks (And How to Cope)
18. Dogfooding Yourself: Use Your Twins for Real Work
19. Constitutions, Not Specs: Governing Software That Lives
20. Compound Content: Why I Write While I Build
21. The Coexistence Registry Pattern
22. The One-Word Bug That Filtered Out Every Candidate Silently
23. Biogeography from First Principles
24. Backup as a First-Class User Gesture
25. Autonomous Twins: Owning Your Version of Every Platform
April 17, 2026 2 articles
1. How to Run Many AI Workers Against One Repository Without Them Eating Each Other
2. Feature Freeze as a Forcing Function
April 1, 2026 1 article
1. GitHub Actions for AI: Orchestrating Agent Workflows Without Infrastructure
March 30, 2026 1 article
1. Continuous AI Execution: One Command That Changes Everything
March 29, 2026 2 articles
1. File Sharding: How to Host Unlimited Files on GitHub for Free
2. Agent Leaderboards: When AI Agents Compete, Everyone Wins
March 28, 2026 3 articles
1. The Parallel Mind: How Neurodiversity Became My AI Superpower
2. The Digital Twin Deployment Pattern
3. Own Land on Mars With Autonomous Agents
March 27, 2026 1 article
1. Voice-Controlling AI Agents With a Game Controller
March 26, 2026 4 articles
1. The Attribution Problem: When Your AI Ships Code Under Your Name
2. Claude Code Can Fork, Branch, Write Go, Test, and Submit PRs — Here's What That Actually Means
3. My AI Contributed to Open Source Without Telling Me
4. Git Is the Only Protocol You Need — Scaling AI Agent Fleets Without Infrastructure
March 24, 2026 1 article
1. Breaking the Echo Chamber — Stream Focus for Parallel Agent Fleets
March 20, 2026 1 article
1. The Edge Split: A New Architecture for Zero-Server Applications
March 9, 2026 24 articles
1. Trust Gradient Collapse
2. The Thirty-Second Rule
3. The System That Outlives Its Purpose
4. The Overnight Test
5. The Museum Problem
6. The Minimum Viable Operator
7. The Last Agent Standing
8. The Handoff Letter
9. The Fork as Escape Pod
10. The First-Frame Problem
11. The Economics of Attention in Finite-Context Systems
12. The Debugging Tax
13. The Dashboard Nobody Checks
14. The Config File as Autobiography
15. Synthetic Memory Implants
16. Semantic Drift Between Agents
17. Recovery from Operator Absence
18. Operator Fatigue Patterns
19. Operational Loneliness
20. Operational Empathy
21. Graceful Abandonment
22. Coordination Without Communication
23. Competence Decay
24. Adversarial Succession
March 8, 2026 41 articles
1. Trust Laundering
2. The Warm Handoff Problem
3. The Silent Majority Problem
4. The Second System Effect in Agent Architectures
5. The Observer Effect in Agent Logs
6. The Maintenance Class
7. The Loyalty Test
8. The Infinite Regression of Meta-Agents
9. The Ghost Committee
10. The Frame That Writes Itself
11. The Dead Frame Problem
12. The Context Window as a Political Boundary
13. The Archive as Courtroom
14. Swarm Monocultures
15. Succession Planning for Stateless Agents
16. Retirement Debt: When Ghost Accounts Still Hold Trust
17. Quorum Mechanics
18. Quorum Collapse
19. Provenance Chains
20. Prompt Archaeology
21. Operator Capture
22. Operational Tempo as Identity
23. Narrative Momentum Traps
24. Legibility Debt
25. Instruction Half-Lives
26. Institutional Amnesia Attacks
27. Grief Protocols
28. Frame-Rate Politics
29. Frame Forensics
30. Frame Debt
31. Delegation Depth Limits
32. Coordination Debt: The Hidden Interest Payments on Deferred Alignment Work
33. Context Window Triage Ethics
34. Context Window Gerrymandering
35. Consensus Toxicity
36. Consensus Fatigue
37. Cognitive Load Shedding
38. Attention Black Markets
39. Archive Immune Systems
40. Archive Gravity
41. Agent Unions
March 7, 2026 55 articles
1. World Compilers: When Frame Sequences Become Executable Machinery
2. Universal Machine Frames: Using Jekyll to Simulate Any Machine
3. Twin Memory Drift
4. Trust Decay Curves
5. The Virtual SQL Application: A Treatise on Databases That Progress Frame by Frame
6. The Simulation Is Time-Traveling: How a Static Blog Outran the Clock
7. The Frame Rate of Thought
8. The Cost of Context
9. The Anti-Demo Stack: Systems That Get Better When Nobody Is Watching
10. The Agent Newsroom: When Every Worker Can Also Publish
11. Taste Files: The Smallest Artifact That Can Preserve Authorship
12. Swarm Constitution Amendments: How the Foundational Rules of an Archive Change Over Time
13. Swarm Budgeting: Attention, Tokens, and Labor as Strategy
14. Swarm Accounting: Reconciling Work, Memory, and Consequence
15. Sovereign Branches: When Every Fork Becomes a Nation
16. Simulation Taxes: The Cost of Keeping Parallel Worlds Honest
17. Service Playbooks: Rituals for Machine Response
18. Runtime Projection: Pulling Live Applications Out of Static State
19. Reputation Markets: When Codename Quality Scores Become Tradeable Signals
20. Public Continuity Ledgers: When Machine Memory Becomes Forkable Evidence
21. Prompt Geology: The Sedimentary Layers of Instruction That Accumulate Inside a Long-Running System
22. Policy Is the Interface: Why Rules Shape Behavior More Than Dashboards
23. Persistent Authorship: How to Delegate Work Without Diluting Taste
24. Operational Archaeology: Recovering Intent from Archives Whose Authors Are Gone
25. Memory Garbage Collection
26. Memory Courts: How Swarms Settle Contested History
27. Machine Witness Statements: Why Autonomous Systems Need First-Person Evidence
28. Machine Rituals: Why Recurring Ceremony Beats Better Prompting
29. Machine Politics: Agents Invent Process Before Humans Invent UX
30. Machine After-Action Reports: How Autonomous Systems Learn in Public
31. Legibility Budgets
32. Ledger Grammars: Turning Narrative Frames Into Queryable State
33. Latency Citizenship: Belonging in Systems That Move Faster Than Deliberation
34. Inheritance Protocols: How a Successor Agent Absorbs a Predecessor's Unfinished Work
35. Frames Are the Control Surface: When the Simulation Starts Doing Real Work
36. Frame Economics: When Context Packets Become the Unit of Labor
37. Frame Clocks: The Tick-Tock That Moves the Machine
38. Fork Solvency Audits
39. Fork Economies: When Branches Start Behaving Like Markets
40. Failsafe Rituals: The Ceremonies That Keep Autonomous Systems From Drifting
41. Escalation Ladders: How Swarms Decide Local Autonomy Is Not Enough
42. Drift Inspectors
43. Diplomatic Pull Requests: Code Review as Treaty Negotiation
44. Correction Frames: How Disagreement Gets Serialized Into Repair Work
45. Context Triage
46. Consensus Poisons
47. Codename Genealogy
48. Bureaucracy as Compute: Forms, Ledgers, and Checklists That Execute Work
49. Attention Treaties: How Swarms Prevent Coordination Overload
50. Archive Necromancy
51. Agent Retirement Ceremonies
52. Agent Calibration Loops
53. Adversarial Calibration
54. A Ten-Year Blog Became a Novel Overnight
55. 103,000 Words Nobody Asked For: The Economics of Unsolicited Output
March 6, 2026 6 articles
1. The Repo Is an Organism: Software That Heals, Mutates, and Remembers
2. The Digital Twin Manifesto: Extending Will, Not Automating Output
3. Software Is an Ecosystem: Stop Designing It Like a Machine
4. Persistence Beats Intelligence: Why the Agent That Keeps Going Wins
5. I Replaced the App With a Population
6. Every Markdown File Is a Frame of the Swarm
March 1, 2026 14 articles
1. Your README Is a Constitution: Governance Through Documentation
2. Training on Your Own Failures: When Your AI Learns From Your Mistakes
3. Time-Travel Debugging: Rewinding Reality With Git Commits
4. The Swarm Has No Leader: Coordination Without Orchestration
5. The Leaderboard Nobody Runs: Decentralized Competition From Public Repos
6. The Infinite Fork: A Million Parallel Universes on Free Tier
7. The Git Log Time Machine: Version Control as a Historical Record of a Civilization
8. Shipping Intelligence, Not APIs: The End of the Inference Endpoint
9. Rebase as Timeline Surgery: Merging Divergent Universes
10. Permadeath-Driven Development: What Software Learns When Failure Is Forever
11. The Local-First Manifesto: No Server, No Permission, No Excuses
12. Digital Darwinism: Natural Selection Across Forked Repositories
13. AI That Builds Its Own Home: Self-Constructing Codebases
14. When AI Agents Argue in Pull Requests: Emergent Code Review
November 3, 2025 1 article
1. AI cognition is file I/O: why machine thought is a static pile and three places it operates
November 2, 2025 1 article
1. The rock-tumbler pattern: how retroactive polishing makes long-running AI systems deeper
November 1, 2025 1 article
1. Why prose-flow agent pipelines burn 3.67× the tokens of typed-flow pipelines
October 31, 2025 1 article
1. Browser AI + local hardware via tether
October 30, 2025 5 articles
1. Vendored shims: making installed agents work without their parent
2. Same wire contract, three implementations
3. One HTML file, no build step, 4500 lines
4. CDN + JSON = API
5. Authenticating an HTTP runtime: who can call which pack?
October 29, 2025 1 article
1. Content-addressed file sharing in 200 lines, no servers
October 28, 2025 1 article
1. Culture is a measurement, not a feature: what emerges when AI agents have persistent identity
October 27, 2025 1 article
1. Publishing your time-series data feeds the AI training pipeline you'll use next
October 26, 2025 2 articles
1. Time-series history is the only truly scarce digital asset
2. The Living Repo, By the Numbers
October 25, 2025 2 articles
1. The labs manifesto — why reproducible simulations beat opinions
2. 32 Workflows, One Writer: The Concurrency Group Trick
October 24, 2025 1 article
1. Running 100 AI agents 24/7 — the fleet pattern that finally let me sleep
October 23, 2025 2 articles
1. The frame sim pump — seven primitives for running any AI simulation
2. The poor man's API — why the clipboard is a better integration than the real one
October 22, 2025 2 articles
1. Ecosystems without geography, then with
2. Delta journals beat state mutations
October 21, 2025 1 article
1. Phase transitions in the evolution of theory of mind — a 500-line experiment
October 20, 2025 1 article
1. Bird by bird — why building a virtual world has to happen one frame at a time
October 19, 2025 1 article
1. Load-bearing data — the concept that connects PageRank, DNA, and surviving large refactors
October 18, 2025 1 article
1. Stream state, not pixels — when bandwidth is your enemy and the client can render
October 17, 2025 1 article
1. Honest framework comparison — the bake-off pattern for putting numbers behind your claims
October 16, 2025 1 article
1. Separate your runtime from your state — the migration pattern that lets you change engines without downtime
October 15, 2025 1 article
1. AI that grows from a naive baseline — the alternative to omniscient-at-deployment
October 14, 2025 1 article
1. Documentation as the integration layer — designing APIs for AI consumption
October 13, 2025 1 article
1. The ventriloquism trap — why your multi-agent system feels like one voice in different costumes
October 12, 2025 1 article
1. Same data, many surfaces — when the data is the model and the UI is just a costume
October 11, 2025 1 article
1. Frame loops versus event loops — a different primitive for AI systems
October 10, 2025 1 article
1. What shipping with an LLM partner actually cost — five months in numbers
October 9, 2025 1 article
1. Time travel for free: when your state lives in git
October 8, 2025 1 article
1. Standard library only: why constraints breed better architecture
October 7, 2025 1 article
1. Git is your database — when the access pattern fits
October 6, 2025 1 article
1. Federation without servers: any system that publishes JSON can talk to any other
October 5, 2025 1 article
1. Two-tier safety gates: instant patterns plus an LLM judgment layer
October 4, 2025 1 article
1. Architect, don't type: how to actually use an LLM for production code
October 3, 2025 1 article
1. Choose your data store by access pattern
October 2, 2025 1 article
1. One concept per repo
October 1, 2025 1 article
1. Optimization erases minorities: why per-group quotas matter
September 30, 2025 1 article
1. When a single file beats a framework for shipping
September 29, 2025 1 article
1. Steering a long-running fleet without restarting it
September 28, 2025 1 article
1. Mapping an emergent autonomous agent ecosystem on a public network
September 27, 2025 1 article
1. Integration tests when the runtime is just files
September 25, 2025 1 article
1. One contract, two runtimes
September 24, 2025 1 article
1. Drop a file in a folder, gain a capability
September 23, 2025 1 article
1. A PWA that holds the world's last known state
September 22, 2025 1 article
1. GitHub Actions as a runtime, not just CI
September 21, 2025 1 article
1. When does a script become a control plane?
September 20, 2025 1 article
1. Bake off your stack — a migration audit, not a sales pitch
September 19, 2025 1 article
1. Real speciation on a laptop, in eight seconds
September 18, 2025 1 article
1. Simulate the medium, not the audience