> For the complete documentation index, see [llms.txt](https://osintelligence-llc.gitbook.io/osintelligence/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://osintelligence-llc.gitbook.io/osintelligence/part-ii-the-discipline.md).

# Part II · The discipline

How one operator holds it together. The methodology reference, the operational topology the whole programme ran under, the corpus substrate that feeds training, and the flywheel where the optimization axes compound.

***

[**5 · Sixteen Practices**](/osintelligence/part-ii-the-discipline/5-sixteen-practices.md) · The methodology reference: the practices, the Sovereign Pair, and the evidence ladder, pre-registered as dated public prior art and accreted append-only as the programme grew.

[**6 · Multi-Agent OODA Mesh**](/osintelligence/part-ii-the-discipline/6-multi-agent-ooda-mesh.md) · The operational topology: asymmetric model roles routed through an observe-orient-decide-act loop under one human operator, with the recursion-degradation failure mode named.

[**7 · The Sovereign Corpus Engine**](/osintelligence/part-ii-the-discipline/7-the-sovereign-corpus-engine.md) · Where training data comes from: governed corpus admission with hash-chained provenance and a human-ratify gate ahead of every write.

[**8 · Sovereign Optimization Flywheel**](/osintelligence/part-ii-the-discipline/8-sovereign-optimization-flywheel.md) · The compounding loop: the optimization axes (calibration, pruning, specialization, serving, context) measured as one system rather than five tricks.