# Todd Paul Brown Jr. > At the frontier of building with AI agents: production software, shipped in plain English, with the receipts. Personal site. Builds carry an exact status word (built, live, demo, shipped, sold, in development); every figure on /receipts/ names its source and whether it was measured or estimated. ## Pages - [About](https://toddpaulbrownjr.com/about/index.md): Todd Paul Brown Jr.: from a web host's billing team to SEO, marketing operations and Adroit, and now building production software by directing AI agents. - [Builds](https://toddpaulbrownjr.com/builds/index.md): What Todd Paul Brown Jr. has built with AI agents, each marked honestly: built, live, demo, shipped, sold, or still in development. Receipts included. - [Earlier work, 2012–2024](https://toddpaulbrownjr.com/portfolio/index.md): Marketing and operations work by Todd Paul Brown Jr. from 2012 to 2024: brand guides, case studies, websites, dashboards and the tooling behind them. - [Now](https://toddpaulbrownjr.com/now/index.md): What Todd Paul Brown Jr. is building in September 2026: Sadyr's meeting booking, this site's rebuild, a queued LinkedIn run, and a game under playtest. - [Privacy](https://toddpaulbrownjr.com/privacy-policy/index.md): What toddpaulbrownjr.com collects when you visit (anonymous traffic statistics and standard server logs), why, and how to reach me about it. - [Receipts](https://toddpaulbrownjr.com/receipts/index.md): Every headline number on this site, with where it comes from, whether it was measured or estimated, and the date it was pulled. Nothing rounded up. - [Todd Paul Brown Jr.](https://toddpaulbrownjr.com/index.md): Todd Paul Brown Jr. builds production software by directing AI agents in plain English, and publishes the receipts: tests, commits, and what broke. - [Writing](https://toddpaulbrownjr.com/writing/index.md): Long-form write-ups on building production software with AI agents: what happened, the evidence, and what broke. Often published here before LinkedIn. ## Builds - [LEGION](https://toddpaulbrownjr.com/builds/legion/index.md): LEGION is the multi-agent operating system Todd runs Adroit on: Claude Code sessions doing real work inside tiered rules the tooling itself enforces. - [Content Ops](https://toddpaulbrownjr.com/builds/content-ops/index.md): Content Ops is Adroit's live, evidence-first content pipeline: discovery, knowledge base and drafts, with a human approving each change. 896 tests passing. - [Sadyr](https://toddpaulbrownjr.com/builds/sadyr/index.md): Sadyr is Adroit's revenue agent: it answers only from approved sources and supports the sales rep. Live platform, two working demos, no paying customers. - [A game in development](https://toddpaulbrownjr.com/builds/game-in-development/index.md): An action RPG in Unreal Engine, built with Claude Code: 81,707 lines of gameplay C++, a playable build under playtest, not shipped. The numbers underneath. - [Cadence](https://toddpaulbrownjr.com/builds/cadence/index.md): Cadence is a fully local proofread-by-ear app Todd built because he is dyslexic. 368 tests, 6.94 hours of narration, zero failed segments. Shipped to one. - [Brain Vault](https://toddpaulbrownjr.com/builds/brain-vault/index.md): The Brain Vault is Todd's ADHD accessibility system: a local-first knowledge base an AI agent can navigate. The architecture and numbers, never contents. - [Earlier work, 2012–2024](https://toddpaulbrownjr.com/builds/earlier-work/index.md): Todd Paul Brown Jr.'s marketing and operations work, 2012 to 2024: brand guides, case studies, websites and dashboards. The years today's work grew from. ## Writing - [A model is not a product: the harness I built on the fly](https://toddpaulbrownjr.com/writing/a-model-is-not-a-product/index.md): The speech model was the easy part. Chunking, retries, sync, caching and a UI I trust are what made Cadence a product I use every day. - [A revenue agent that answers only from what you approved](https://toddpaulbrownjr.com/writing/a-revenue-agent-that-answers-from-what-you-approved/index.md): Sadyr answers pre-sales questions from approved sources only, abstains when it isn't sure, and hands the conversation to a person. - [A review agent found two bugs that would have made a calm-audio app secretly stressful](https://toddpaulbrownjr.com/writing/the-pan-law-that-cancelled-itself/index.md): ** A review agent found a self-canceling pan law and a fake limiter in a generative audio app; listening alone would likely never have isolated either. - [A rule it can argue with: where rules for agents belong](https://toddpaulbrownjr.com/writing/a-rule-it-can-argue-with/index.md): A rule in a prompt gets broken; a rule in the tooling gets enforced. The heredoc hook that blocked the third violation and the lesson another builder can use. - [A unit test deleted a real 4.7 GB model file](https://toddpaulbrownjr.com/writing/a-test-deleted-a-real-model/index.md): A subagent-written test undid its own sandbox and deleted a real 4.7 GB model; the tool's own failure log caught it and showed why recovery worked. - [A website built to be read by machines as well as people](https://toddpaulbrownjr.com/writing/a-website-for-machines-and-people/index.md): Adroit's website rebuilt with a machine-readable layer: 74 pages updated, 17 created, JS from 103 KB to 2 KB, Markdown twins and llms.txt live. - [ADHD, novelty, and building the right room](https://toddpaulbrownjr.com/writing/adhd-novelty-and-building-the-right-room/index.md): Novelty-seeking isn't a deficit if you build the environment where it's fuel. How I built systems and tools that work with ADHD, not against it. - [Adversarial audits left in the code](https://toddpaulbrownjr.com/writing/audits-left-in-the-code/index.md): ** Three severity-coded adversarial audit rounds live as comments in shipped code; one field-coverage audit caught a form field silently dropped for two months. - [Aligning incentives, then and now: incentives for agents are permissions](https://toddpaulbrownjr.com/writing/aligning-incentives/index.md): People respond to what they are scored on; agents respond to what they are allowed to do. Aligning either one is designing the scoreboard. - [An agent-operated job-search pipeline with two Python packages and no database](https://toddpaulbrownjr.com/writing/an-agent-operated-job-search-pipeline/index.md): ** A job-search pipeline with 13-dimension scoring, file-based state, and an agent-facing API of pure re-exports — built with two Python packages and no… - [An on-device interactive fiction engine: 38 decisions, 1,010 tests, and why it's parked](https://toddpaulbrownjr.com/writing/inside-the-interactive-fiction-engine/index.md): ** A parked on-device interactive fiction engine: 1,010 tests, 38 decisions, zero network calls, and what I learned building the harness that could refuse… - [Audio Apothecary: a generative audio lab with an honesty policy](https://toddpaulbrownjr.com/writing/inside-audio-apothecary/index.md): A generative audio app built in six days with 1,211 tests — and a review agent that found the showstopper bugs the test suite missed. - [Billing, migrations and my first team](https://toddpaulbrownjr.com/writing/billing-migrations-and-my-first-team/index.md): ** Hired into Bluehost billing to absorb two acquired hosts, I learned fast, led training and escalations, and walked out knowing WordPress well enough to… - [Built it, piloted it, deleted it](https://toddpaulbrownjr.com/writing/built-it-piloted-it-deleted-it/index.md): I built an orchestration platform, piloted it successfully, then deleted it because plain Python and JSON files handled the workflow with less complexity. - [Built, live, demo, shipped, sold](https://toddpaulbrownjr.com/writing/five-status-words/index.md): Agents made 'built' cheap, so the status word is now the claim. Five words the industry uses as one, and a real example of each from Todd's own work. - [Cadence: the proofreading tool I built because my brain is worst at proofreading](https://toddpaulbrownjr.com/writing/inside-cadence/index.md): A local proofreading app that reads Markdown aloud with synced highlighting. 368 tests, 2,995 segments rendered, zero failed. Built because dyslexia. - [Connect your data: seven systems into one, with nothing rolled back](https://toddpaulbrownjr.com/writing/connect-your-data/index.md): Six weeks to connect seven systems for a national sports league. The data mirror is live, additive-only, and leaves the audits in the code. - [Content Ops: an evidence-first content engine, and what it took to run it at scale](https://toddpaulbrownjr.com/writing/inside-content-ops/index.md): Content Ops drafts only from an approved knowledge base. Running it at scale meant 92 of 95 pages in two days, a same-night incident loop, and choosing… - [Curiosity as a catalyst, then and now: I learned it on a toy](https://toddpaulbrownjr.com/writing/curiosity-as-a-catalyst/index.md): ** The same curiosity that took a billing lead into SEO took a toy game's habits into LEGION — neither was a planned pipeline, both times it worked. - [Data at the helm, then and now: the Brown Score, and an audit log with one entry for thirty-one jobs](https://toddpaulbrownjr.com/writing/data-at-the-helm/index.md): A usage score drove client retention at Banyan; an agent that never logged its own actions told nobody anything. Data steers only when it's written down and… - [Deterministic code decides when to speak; the model only chooses the words](https://toddpaulbrownjr.com/writing/code-decides-when-the-model-chooses-words/index.md): Lessons from a failed on-device assistant: put decision logic in code, keep agent memory separate, and watch for SDK instances that don't talk. - [Eight months of not building LEGION](https://toddpaulbrownjr.com/writing/eight-months-of-not-building-legion/index.md): ** I tried to build LEGION three times in 2025 and failed each time. The idea didn't change — what was missing was operator skill and model capability. - [Every layer of computing was a plain-language layer](https://toddpaulbrownjr.com/writing/every-layer-was-a-plain-language-layer/index.md): Each computing layer let people say more in words closer to their own. English prompts are the next layer, and like every layer before, they moved work up. - [Every pitfall I pre-wrote a rule for, my agent still needed reminding about](https://toddpaulbrownjr.com/writing/every-pitfall-i-pre-wrote/index.md): ** Ten documented pitfalls, four real bugs caught in audit, six problems the agent found itself — and the lesson that a rule an agent remembers is just a… - [Faster, lighter, bigger: every model swap I rejected, and the metric that decided](https://toddpaulbrownjr.com/writing/every-model-swap-i-rejected/index.md): Three faster, lighter models lost on prose quality; a bigger model that won on quality was rejected on speed. The decision rule matters more than the… - [Four patch points into a system we had never seen](https://toddpaulbrownjr.com/writing/four-patch-points/index.md): Reverse-engineer an unfamiliar system, then wire a CRM in through four additive patch points with under 80 lines changed and zero edits to existing logic. - [From COO to Chief Innovation Officer: how operations work turned into AI work](https://toddpaulbrownjr.com/writing/from-coo-to-chief-innovation-officer/index.md): ** I joined Adroit as COO in 2021, wrote down every process, then spent years building AI into our workflows until the title changed to Chief Innovation… - [Front-line expertise, then and now: the review agent found what the builders missed](https://toddpaulbrownjr.com/writing/front-line-expertise/index.md): Weekly trainings at NP Accel and an adversarial review agent are the same mechanism: a channel for front-line expertise with no stigma for being wrong. - [Governance for agents that do real things](https://toddpaulbrownjr.com/writing/governance-for-agents/index.md): Once agents can deploy and spend, governance is four rules enforced in tooling: observe freely, mutate with a flag, some actions human-only, no repo… - [Guiding transformation, then and now: seventy-four pages rewritten, staged, cut over](https://toddpaulbrownjr.com/writing/guiding-transformation/index.md): ** A rebrand and a 74-page site rewrite share the same shape: stage everything, verify each piece, cut over only when the check is done. - [I built a tool to audit my own AI claims. It made my numbers smaller.](https://toddpaulbrownjr.com/writing/i-audited-my-own-ai-claims/index.md): I audited my own AI usage claims with a purpose-built collector and four verification passes. The audit cut a 10-billion-token headline to 36 million and… - [I made the agent attack its own diff engine before I trusted it](https://toddpaulbrownjr.com/writing/attack-the-diff-engine/index.md): ** I attacked Cadence's diff engine with adversarial tests before trusting it. Three bugs, three fixes, 368 tests. Why hash-based reconciliation matters. - [Inside LEGION: running agents like an engineering org](https://toddpaulbrownjr.com/writing/inside-legion/index.md): LEGION is the control layer I run Adroit on — a generated state panel, human-triggered sessions, capability packages and agent governance, built after eight… - [Inside the Brain Vault: layers, triage and seventeen playbooks](https://toddpaulbrownjr.com/writing/inside-the-brain-vault/index.md): ** Raw and synthesized layers, signal triage, dated confidence labels, PageRank cross-checked, and seventeen agent playbooks instead of vibes. - [Make the agent argue first: plan the plan, defend the design](https://toddpaulbrownjr.com/writing/make-the-agent-argue-first/index.md): Before Cadence got code, a 1,105-line prompt made the agent plan its plan, and a design agent defended 19 features; arguing up front was cheaper than rework. - [Measuring success, then and now: green tests are not a result](https://toddpaulbrownjr.com/writing/measuring-success/index.md): Self-reported evals at chance level and seven passing security probes on an abandoned product: why green tests don't prove you built something worth having. - [Metrics that misled: the flat traffic that wasn't](https://toddpaulbrownjr.com/writing/metrics-that-misled/index.md): A client's flat traffic hid growth and a conversion problem. "Why now?" is the question that separates a metric from a story. - [Milkman: a benchmarking tool that refuses downloads that won't fit](https://toddpaulbrownjr.com/writing/inside-milkman/index.md): A local-model benchmarking tool built in 31.4 hours that tunes itself to hardware, refuses downloads that won't fit, and logs its own failures. - [My listening log became the bug tracker](https://toddpaulbrownjr.com/writing/my-listening-log-became-the-bug-tracker/index.md): ** A plain-language listening log sat directly above an agent's dated root-cause entries, and one 'crackling' report turned out not to be a bug. - [My READMEs lied about my tests](https://toddpaulbrownjr.com/writing/my-readmes-lied-about-my-tests/index.md): Every stale test count Todd found this year undercounted, across three projects and three stacks. Why docs drift one way, and why agents make it worse. - [My second brain is an ADHD accessibility system an agent can navigate](https://toddpaulbrownjr.com/writing/my-second-brain-is-an-accessibility-system/index.md): The Brain Vault: 1,765 documents in two layers, 50 pillars, 17 playbooks. An ADHD accessibility system designed so an agent can navigate it and act on real… - [One platform, two opposite audiences](https://toddpaulbrownjr.com/writing/one-platform-two-audiences/index.md): Two working Sadyr demos — public insurance explainer and gated sports rulebook assistant — run on one platform with opposite audiences and no re-architecting. - [Operator or agent: make authorship a field](https://toddpaulbrownjr.com/writing/operator-or-agent/index.md): The one architectural decision an agent made without permission was caught because every decision record now says who made it; authorship belongs in the data. - [Own the center; wrap, rent or swap the edges](https://toddpaulbrownjr.com/writing/own-the-center/index.md): Sadyr's architecture rule from a production incident: own the policy and logic that make your system yours; treat models, CRMs and channels as swappable edges. - [Parallel is a budget, not a speed setting](https://toddpaulbrownjr.com/writing/parallel-is-a-budget/index.md): ** Ten parallel subagents burned a five-hour budget in 33 minutes. The fix was a cap and a rule, not a bigger plan — and the same lesson applies to… - [Pre-defined metrics, then and now: my eval measured nothing](https://toddpaulbrownjr.com/writing/pre-defined-metrics/index.md): Define success before the work starts and verify the measurement actually tests what matters, or your eval will agree with whatever the model says. - [Sadyr: blueprint to production in twelve days, and what it answers from](https://toddpaulbrownjr.com/writing/inside-sadyr/index.md): ** Sadyr went from blueprint to production in twelve days. Here's the architecture, the bugs, the numbers, and what it answers from. - [Seven security probes passed. The product still failed.](https://toddpaulbrownjr.com/writing/seven-probes-passed-the-product-failed/index.md): ** An on-device assistant passed seven security checks and then sat unused for 55 days, because it only delivered value when I babysat it. - [Seven systems, about 2,400 teams, six weeks, zero rollbacks](https://toddpaulbrownjr.com/writing/seven-systems-six-weeks-zero-rollbacks/index.md): ** A revenue-operations rebuild for a national amateur sports league: seven systems into one CRM, additive-only changes, and no rollbacks. - [Shipping production software in plain English: what it actually takes](https://toddpaulbrownjr.com/writing/shipping-software-in-plain-english/index.md): What it actually takes to ship production software in plain English: the gates, the status words, the failures, and the parts that still need a human. - [Six shipped builds ran settings no eval had validated](https://toddpaulbrownjr.com/writing/evals-passed-the-wrong-app/index.md): Six Android builds of an on-device fiction engine ran sampler settings the eval suite never tested, because the runtime silently filled defaults. - [Stalled vs blocked](https://toddpaulbrownjr.com/writing/stalled-vs-blocked/index.md): A project that goes quiet is either stalled or blocked. The game's 63-day gap was the second kind, and knowing the difference changed every build after it. - [Start from someone else's repo](https://toddpaulbrownjr.com/writing/start-from-someone-elses-repo/index.md): I cloned open-source repos and built on top. Credit upstream, keep changes additive, and know which bugs are theirs — those are the rules. - [Status is derived, not reported](https://toddpaulbrownjr.com/writing/status-is-derived-not-reported/index.md): Docs drift and agents trust them. The fix: one generated page, status computed from what exists, not what was last written down. - [The 'I told you so' board: taking the stigma out of being wrong](https://toddpaulbrownjr.com/writing/the-i-told-you-so-board/index.md): Teams hide mistakes when being wrong is costly. A board that celebrates testing ideas made failure normal—and it's why I publish mine now. - [The answer gate that divided confidence by word count](https://toddpaulbrownjr.com/writing/the-answer-gate-that-divided-by-word-count/index.md): Sadyr's confidence gate divided retrieval scores by word count; every test passed, every real question abstained. Found and fixed in one day. - [The art of persuasion, then and now: the churn warning nobody acted on, and the check nobody built](https://toddpaulbrownjr.com/writing/the-art-of-persuasion/index.md): A churn warning ignored in 2016 and a values check nobody built in 2026: both times the information existed, and both times the system wasn't built to act… - [The creature-battler where my working habits started](https://toddpaulbrownjr.com/writing/the-creature-battler/index.md): A parked Godot toy that taught me battle math in a spreadsheet first, session timers before my business had them, and that tests don't replace play. - [The handoff file: every session rewrites it, none appends](https://toddpaulbrownjr.com/writing/the-handoff-file/index.md): Agents forget between sessions. A single handoff file, rewritten not appended, plus a stand-up and wind-down ritual keeps long projects coherent. - [The infrastructure behind a game in development](https://toddpaulbrownjr.com/writing/the-infrastructure-behind-a-game-in-development/index.md): ** 81,707 lines of C++, 519 subagent conversations, 250 logged decisions: the measured infrastructure behind a game in development built with Claude Code. - [The layer finally arrived in my native language](https://toddpaulbrownjr.com/writing/the-layer-arrived-in-my-native-language/index.md): I'm dyslexic and proofreading by eye is the task my brain is worst at, so I built Cadence to let me proofread by ear instead — the layer thesis made personal. - [The overnight run, and the gate that earned it](https://toddpaulbrownjr.com/writing/the-overnight-run/index.md): Three engine systems landed overnight with nobody at the keyboard. The overnight part is not the story. The gate that could refuse the result is. - [Three production defects, found and fixed the same night](https://toddpaulbrownjr.com/writing/three-defects-the-same-night/index.md): ** Three production defects in Content Ops — a runs race, an artifacts race, and a polling fan-out — surfaced and fixed the same night from structured logs. - [Thriving in chaos, then and now: a backlog of thousands, and a second brain built for how I think](https://toddpaulbrownjr.com/writing/thriving-in-chaos/index.md): ** Chaos is usually a system that depends on someone remembering. The fix at 180Fusion and the fix in the Brain Vault both write the memory down where it… - [Two bugs that weren't mine: a 30-year-old Windows quirk and an upstream voice bug](https://toddpaulbrownjr.com/writing/two-bugs-that-werent-mine/index.md): ** A dyslexia proofreading app broke on a 30-year-old Windows quirk and an upstream voice bug; building on other people's code means debugging it too. - [What Claude Code's own report says about how I work](https://toddpaulbrownjr.com/writing/what-claude-codes-report-says-about-how-i-work/index.md): Claude Code's Insights report on eight weeks of my sessions: 855 messages, 262 hours, 195 commits. What it got right, what goes wrong, and what I changed. - [What I actually do when the agent writes the code](https://toddpaulbrownjr.com/writing/what-i-do-when-the-agent-writes-the-code/index.md): The agent writes the code and tests. I direct, decide, and validate. 157 prompts in two weeks when 75% of the C++ was written—measured. - [Who Adroit is for: businesses competing against bigger money](https://toddpaulbrownjr.com/writing/who-adroit-is-for/index.md): ** Adroit is a Growth Partner for businesses competing against bigger money, running on the same systems it sells: revenue operations, marketing… - [Your documentation is source code now](https://toddpaulbrownjr.com/writing/your-documentation-is-source-code/index.md): ** About 40% of the game's 104,912 documentation lines exist so the agent can read itself back into context — measured proof that docs are source code now. - [Your graph view is a hairball because it treats every link as equal](https://toddpaulbrownjr.com/writing/your-graph-view-is-a-hairball/index.md): Sparse top-K edges, PageRank as mass and a seeded layout turn a knowledge graph into a map. A nebula that never rendered showed what 31 tests missed. ## The Anatomy of Corruption and Corrigibility (field guide) - [The Anatomy of Corruption and Corrigibility](https://toddpaulbrownjr.com/corrigibility/index.md): A field guide to diagnosing corruption and failure modes in information systems: what a system is really optimizing, and whether it can still be corrected. - [Stated Goals and Revealed Goals](https://toddpaulbrownjr.com/corrigibility/stated-and-revealed-goals/index.md): Every system has a goal it states and a direction it actually moves. How to find the second, and how to judge it without guessing at anyone's soul. - [Systems That Persist as Patterns](https://toddpaulbrownjr.com/corrigibility/systems-that-persist-as-patterns/index.md): An institution outlives everyone in it because it is a pattern, not a roster. What that pattern is made of, and the ways it quietly goes wrong. - [Three Kinds of Together: Grex, Collegium, Egregore](https://toddpaulbrownjr.com/corrigibility/three-kinds-of-together/index.md): Groups hold together in three different ways: local rules, shared standards, and shared belief. Each fails differently and needs a different repair. - [Sucralosis: Reward Without Realization](https://toddpaulbrownjr.com/corrigibility/sucralosis/index.md): When a system chases a stand-in for what it wants, it finds every way to hit the number without the real thing. Three questions that catch it early. - [Cancrity: Growth That Severs Its Feedback](https://toddpaulbrownjr.com/corrigibility/cancrity/index.md): Some systems thrive by using up what they depend on. How to tell that from ordinary competition, and a three-line test that keeps the diagnosis honest. - [Contracorrigibility: How Systems Disable Their Own Correction](https://toddpaulbrownjr.com/corrigibility/contracorrigibility/index.md): Some systems can't see their failures. Others turn their own correction machinery against the fix. How to tell bit rot from a rootkit, and what each needs. - [Anti-Memesis: Why Critique Bounces Off](https://toddpaulbrownjr.com/corrigibility/anti-memesis/index.md): Some critiques are never rejected because they never register. The tell: structural feedback turned into incident-sized feedback. "Give me the list." - [The Telonic Audit](https://toddpaulbrownjr.com/corrigibility/the-telonic-audit/index.md): A repeatable audit: check the system can hear the answer, then ask about the gap, the coupling, and the standing. Who pays for asking? - [Markers and Tests](https://toddpaulbrownjr.com/corrigibility/markers-and-tests/index.md): Signs of health are cheap to fake. The markers worth trusting are the ones only real correction machinery can produce, read now, before the outcome. - [Rates and Forecasts](https://toddpaulbrownjr.com/corrigibility/rates-and-forecasts/index.md): Swap "will it happen?" for "which rate is faster?", state the method's bounds, pin forecasts with a six-slot register, and keep the framework honest. - [AI Systems](https://toddpaulbrownjr.com/corrigibility/ai-systems/index.md): Specification gaming, sycophancy and eval-gaming: the proxy failure in machinery with no self. Two senses of corrigibility, and instruments for builders. - [Public Institutions](https://toddpaulbrownjr.com/corrigibility/public-institutions/index.md): Agency capture, movements that win and have nothing to become, and four requirements for building institutions that keep correcting themselves on purpose. - [Corporations and the Workplace](https://toddpaulbrownjr.com/corrigibility/corporations-and-the-workplace/index.md): The dashboard that ate the mission, status rounding up the chain, and a company going deaf one reorg at a time, with an audit you can run from any seat. - [The Individual Reader](https://toddpaulbrownjr.com/corrigibility/the-individual-reader/index.md): You are also a system: proxies replace their reasons, a life narrows around one source of meaning, and a portfolio of commitments keeps you correctable. - [Coda: The Bounds of the Instrument](https://toddpaulbrownjr.com/corrigibility/coda/index.md): Where the framework is strong and weak, how it could go unfalsifiable, what would prove it wrong, and why it cannot reach a captured reader. - [Glossary](https://toddpaulbrownjr.com/corrigibility/glossary/index.md): The coined terms of The Anatomy of Corruption and Corrigibility, defined operationally: telon, sucralosis, cancrity, contracorrigibility, anti-memesis and more. - [How This Was Written: Debugging a Philosophy](https://toddpaulbrownjr.com/corrigibility/how-this-was-written/index.md): The framework started as my own philosophy. I treated it as code, mapped every claim and its evidence, and debugged it with a coding agent until it held up. - [Changelog](https://toddpaulbrownjr.com/corrigibility/changelog/index.md): What changed between drafts of The Anatomy of Corruption and Corrigibility, and why. Claims get corrected here in the open. ## Elsewhere - [LinkedIn](https://www.linkedin.com/in/toddpbrownjr/): where the shorter versions run - [Adroit](https://adroitmktg.com/): the company Todd builds for ## Optional - [Full corpus](https://toddpaulbrownjr.com/llms-full.txt): every page and article as Markdown, concatenated.