# A builder's codex > A primary-source library of operator insights. Atomic claims, named operators, verifiable sources. The codex is a structured corpus of cross-domain operator wisdom across product, PMM, GTM, AI-native, design, engineering, leadership, sales, growth, research, and founder craft. Every claim traces back to its primary source. Insights carry mechanism, conditions, evidence, signals, counter-evidence, and cross-references. ## Index - [INDEX.json](https://codex.iamkesava.com/insight-library/INDEX.json): Machine-readable index of all insights, operators, patterns, contradictions, playbooks. Use this as the canonical entry point. - [INDEX.md](https://codex.iamkesava.com/insight-library/INDEX.md): Human-readable index. - [latest.json](https://codex.iamkesava.com/insight-library/latest.json): Newest release entry + last 30 daily entries. ## Browse - [Home](https://codex.iamkesava.com/): Tier A claims, domain index. - [All operators](https://codex.iamkesava.com/#/operators) - [Synthesis patterns](https://codex.iamkesava.com/#/patterns): Where 3+ operators converge on the same claim. - [Methodology playbooks](https://codex.iamkesava.com/#/playbooks) - [Timeline](https://codex.iamkesava.com/#/timeline): All insights by source date, density-scaled. - [Release log](https://codex.iamkesava.com/#/today): What's new in the corpus, newest first. ## Schema Every insight card frontmatter: - `id`: globally unique, kebab-case, prefixed `ins_` - `operator`: primary author (full name) - `co_operators`: list of additional named authors - `operator_role`: role at time of source - `source_url`, `source_type`, `source_title`, `source_date`: primary source attribution - `captured_date`: when added to corpus - `domain`: taxonomy domains (product, pmm, gtm, ai-native, …) - `lifecycle`: taxonomy lifecycle tags - `maturity`: foundational | applied | frontier - `artifact_class`: framework | playbook | template | workflow | metric-model | case-study | excerpt - `score`: { originality, specificity, evidence, transferability, source } (1-5 each) - `tier`: A | B | C - `related`: list of related insight ids Body sections: `Claim` (one sentence), `Mechanism` (why it works), `Conditions` (when it applies, when it does not), `Evidence` (quote, example, data), `Signals` (how to know it's working), `Counter-evidence` (where it fails, who disagrees), `Cross-references`. ## Authoring - [Ingest protocol](https://github.com/k3sava/ab-codex/blob/main/insight-library/00_meta/INGEST-PROTOCOL.md) - [Insight card schema](https://github.com/k3sava/ab-codex/blob/main/insight-library/00_meta/insight-card-schema.md) - [Daily release-log contract](https://github.com/k3sava/ab-codex/blob/main/insight-library/daily/README.md) ## Optional - [GitHub repository](https://github.com/k3sava/ab-codex): MIT licensed. Insights are open source; raw sources retain original copyright.