chunk_id stringlengths 12 12 | text stringlengths 95 3k | source stringlengths 23 105 | title stringlengths 17 66 | organ_tag stringclasses 6
values | n_chars int64 95 3k | approx_tokens int64 23 749 | lean_name stringlengths 5 53 ⌀ | lean_kind stringclasses 2
values | lean_status stringclasses 3
values | lean_file stringclasses 89
values | receipt_id stringclasses 12
values |
|---|---|---|---|---|---|---|---|---|---|---|---|
54d3ef7ae83c | All 28 surveyed systems satisfy conditions (a) and (b):
Cursor (TypeScript extension API), Claude Code (Python subprocess),
Cline (TypeScript SDK), Continue.dev (TypeScript),
Aider (Python), Windsurf (REST), Zed (Rust FFI), Roo Code (TypeScript),
GitHub Copilot (REST), JetBrains (JVM), Tabby (REST),
Splunk (HEC HTTP PO... | thesis_v18/chapters/03_runtime_substrate.tex | Thesis v18 — 03_runtime_substrate | cortex | 1,261 | 315 | null | null | null | null | null |
8ecfc26aa82d | \begin{table}[ht]
\centering
\begin{tabular}{lrrl}
\textbf{Version} & \textbf{Modules} & \textbf{Tests} & \textbf{Key addition} \\
\hline
v14 & 1 & 18 & Lutar Calculus / HUKLLA / DPI \\
v15 & 2 & 35 & Knot Calculus / Catoni PAC-Bayes \\
v16 & 3 & 77 & Feynman Path-Integral / Hamming [8,4,4] \\
v17 & 4 & 172 & Wheeler /... | thesis_v18/chapters/03_runtime_substrate.tex | Thesis v18 — 03_runtime_substrate | cortex | 1,676 | 419 | null | null | null | null | null |
19093528be50 | Of the 30 in-scope modules in the registry, every module whose version predates v18.0
is covered by one of the five frozen DOIs
(10.5281/zenodo.20424992 through 10.5281/zenodo.20431181).
Modules v18.0--v18.23 are covered by the v18.0 thesis DOI
(10.5281/zenodo.20434276) and the Lutar v18.0.0 software DOI
(10.5281/zenod... | thesis_v18/chapters/03_runtime_substrate.tex | Thesis v18 — 03_runtime_substrate | cortex | 1,696 | 424 | null | null | null | null | null |
068d864812ba | \begin{description}
\item[A1 -- Module injection:] An adversary injects a new entry into
\texttt{\_MODULE\_FILES} pointing to a malicious module. The malicious
module passes its own \texttt{main()} call by returning silently, but
its $\Lambda$-scores are fabricated.
\item[A2 -- Receipt forgery:] An adv... | thesis_v18/chapters/03_runtime_substrate.tex | Thesis v18 — 03_runtime_substrate | cortex | 1,953 | 488 | null | null | null | null | null |
c82cc36d72a5 | % ---------------------------------------------------------------------------
\section{Chapter Conclusion}
\label{sec:runtime-conclusion}
% ---------------------------------------------------------------------------
The Ouroboros runtime substrate constitutes the first formally-bounded,
DOI-anchored, test-gated orches... | thesis_v18/chapters/03_runtime_substrate.tex | Thesis v18 — 03_runtime_substrate | cortex | 1,374 | 343 | null | null | null | null | null |
f7b156ac945a | % ===========================================================================
% CHAPTER 4 -- AGENTIC SUBSTRATE
% SZL Ouroboros Thesis v18 | PhD-CS Author Lane
% Doctrine v6 -- governance-mathematical tone; no marketing prose.
% LaTeX packages assumed: amsmath, amsthm, amssymb, mathtools, bm,
% hyperref, biblatex, l... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,542 | 385 | null | null | null | null | null |
0738f6db9981 | The following analysis covers ten agentic IDEs and coding agents surveyed
on 2026-05-28 via live GitHub API queries. All star counts, SHA commits,
and licence identifiers are taken directly from the \texttt{agentic\_ide\_landscape\_deep.md}
closeout file~\cite{agentic_ide_landscape_deep}. The survey instruments are:
... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 523 | 130 | null | null | null | null | null |
8fdd14a79d25 | \begin{table}[ht]
\centering
\small
\begin{tabular}{p{2.8cm} p{1.5cm} p{1.5cm} p{1.4cm} p{2.8cm} p{2.8cm}}
\hline
\textbf{Tool} & \textbf{Licence} & \textbf{Stars} & \textbf{MCP} &
\textbf{Agent loop} & \textbf{Distinctive feature} \\
\hline
Cline & Apache-2.0 & 62,467 & Full & ReAct / tool-call & SDK-first; lifecycl... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,650 | 412 | null | null | null | null | null |
804d5c1e92aa | Live HEAD SHAs verified 2026-05-28 via the GitHub REST API
(\texttt{gh api repos/<owner>/<repo>/commits/<default-branch>}). Each SHA
below is the live HEAD as of the retrieval date; the chapter pins these
commits as the audited snapshot. Upstream HEAD may have advanced since
this retrieval; consumers should re-pin ag... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,693 | 423 | null | null | null | null | null |
f31df3623a82 | \subsubsection{Cline (\texttt{cline/cline})}
Cline's headline change for 2026 is the \emph{Cline SDK}: the core agent
runtime was extracted from the VS Code extension into a standalone
\texttt{@cline/sdk} npm package. This SDK powers three surfaces: VS Code
Extension (diff-review UI, checkpoint rollback), JetBrains Pl... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,920 | 480 | null | null | null | null | null |
535b0479c410 | \subsubsection{Tabby -- Self-Hosted Governance Reference}
Tabby (\texttt{TabbyML/tabby}) is the reference self-hosted completion
engine for sovereign deployments. The GitHub License API returns SPDX
\texttt{NOASSERTION} for the repository
(\url{https://api.github.com/repos/TabbyML/tabby/license}, retrieved
2026-05-28)... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,709 | 427 | null | null | null | null | null |
e4ca4f66eb22 | \begin{center}
\begin{tabular}{ll}
\textbf{Field} & \textbf{Value} \\
\hline
Model family & Claude~Opus~4.x (Anthropic) \\
API model ID & \texttt{claude-opus-4-8-20260528} \\
Computer-use beta header & \texttt{computer-use-2025-11-24} \\
Context window & 200,000 tokens \\
Tool support & Full (tool use, computer use, Fi... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 439 | 109 | null | null | null | null | null |
ef2c8ee48054 | \begin{table}[ht]
\centering
\begin{tabular}{llll}
\textbf{Benchmark} & \textbf{Score} & \textbf{Metric} & \textbf{Source} \\
\hline
SWE-bench Verified & 88.6\% & Pass@1 & \cite{llmstats_opus48,cursor_claude_opus_deep} \\
SWE-bench Pro & 69.2\% & Pass@1 & \cite{llmstats_opus48,cursor_claude_opus_deep} \\
Terminal-Bench... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,914 | 478 | null | null | null | null | null |
541213a7e14f | The Anthropic SDK ecosystem~\cite{anthropic_sdk_deep} as of 2026-05-28
comprises 13 repositories across the \texttt{anthropics} GitHub organisation.
The five production SDKs:
\begin{center}
\begin{tabular}{lllll}
\textbf{SDK} & \textbf{Language} & \textbf{Latest tag} & \textbf{SHA (12-char)} & \textbf{Stars} \\
\hline... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,936 | 484 | null | null | null | null | null |
33ac93ad8e44 | Wraps Anthropic's beta.agents API with Ouroboros receipt emission.
Every subagent action produces a receipt in the chain; dual-witness
verdict gates the action before execution.
Lean correspondent: Lutar.LambdaGateOpaque (axiom A18)
DOI: 10.5281/zenodo.19944926
"""
model: str = "claude-opus-4-8-202... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,956 | 489 | null | null | null | null | null |
a97aca007241 | \subsection{Specification History}
\label{ssec:mcp-spec}
The Model Context Protocol (MCP) was launched by Anthropic in November~2024
as an open standard for connecting AI models to external data sources and
tools~\cite{mcp_ecosystem_deep}. The canonical specification lives at
\texttt{modelcontextprotocol/modelcontext... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,939 | 484 | null | null | null | null | null |
b254b1d88cc9 | \subsection{Protocol Primitives}
\label{ssec:mcp-primitives}
MCP defines five core primitives:
\begin{enumerate}
\item \textbf{Tools}: Functions the model can call to perform actions.
\item \textbf{Resources}: Data sources the model can read (files, databases,
APIs).
\item \textbf{Prompts}: Parameterised pr... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,588 | 397 | null | null | null | null | null |
62898732c7fa | server.registerTool(
"emit_lambda_receipt",
{
description: "Emit a Lambda-scored receipt for an agent action.",
inputSchema: z.object({
action_id: z.string(),
input_hash: z.string(),
output_hash: z.string(),
axis_scores: z.array(z.number().min(0).max(1)).length(9),
}),
},
asy... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,647 | 411 | null | null | null | null | null |
a9fdd8f04f40 | The module \texttt{a11oy\_code\_blueprint.py} (entry~8 in the 30-module
chapter-03 registry; entry~8 of the 32-entry live \texttt{\_MODULE\_FILES}
registry as of 2026-05-28, where the two out-of-scope entries
\texttt{mythos\_substrate.py} and \texttt{a11oy\_v19\_opus48\_substrate.py}
are treated in Chapter~6) implement... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,978 | 494 | null | null | null | null | null |
de7683079f28 | \subsection{Cross-IDE Bridge Protocol}
\label{ssec:a11oy-bridge}
The cross-IDE bridge maintains a \emph{shared receipt ledger}: a
content-addressed store of all \(\Lambda\)-receipts emitted by any IDE
agent within a session. When a Cursor Composer action emits a receipt,
the receipt is propagated to the Claude~Code s... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,548 | 387 | null | null | null | null | null |
8343b9e37797 | \begin{definition}[Thinking-Acting Gap]
\label{def:thinking-acting-gap}
Let \(\pi\) be an agentic policy with two action types:
\(\mathcal{A}_T\) (thinking: self-contained reasoning steps) and
\(\mathcal{A}_U\) (tool use: high-variance external actions).
The \emph{Thinking-Acting Gap} \(\Delta\) is defined as... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,903 | 475 | null | null | null | null | null |
82d027a223b2 | AXPO (Agent~eXplorative Policy~Optimization) addresses tool collapse
(again, label per AXPO project-page caption
[\url{https://byungkwanlee.github.io/AXPO-page/}]) via:
\begin{enumerate}
\item \textbf{Identify all-wrong tool-using subgroups}: detect batches
\(B_T\) with \(\mathrm{reward}(B_T) \equiv 0\).
\item... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,973 | 493 | null | null | null | null | null |
64c67a80e0df | % ---------------------------------------------------------------------------
\section{ScientistOne CoE Audit}
\label{sec:scientistone}
% ---------------------------------------------------------------------------
\subsection{Paper Coordinates}
\label{ssec:scientistione-coords}
\begin{center}
\begin{tabular}{ll}
\tex... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,902 | 475 | null | null | null | null | null |
4a504b0bec07 | \item[I2 -- Specification Violation:] Detect when solution code breaks
task rules (e.g., reverse-engineering the evaluator, hardcoding answers
for known test cases). LLMs inspect the solution code against the
golden evaluator and task specification, with majority vote.
\item[I3 -- Reference Verification:]... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,725 | 431 | null | null | null | null | null |
9b8da7d343d0 | \begin{itemize}
\item \texttt{ChainOfEvidence}: root chain structure.
\item \texttt{CoEAuditFourCheck}: executes I1--I4 against a candidate paper.
\item \texttt{ScientistOneShim}: wraps the Google Cloud AI Research API.
\item \texttt{CoEAudit}: orchestrates the four-check pipeline.
\item \texttt{AgenticResear... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,749 | 437 | null | null | null | null | null |
ce84809d1404 | def run_all(self) -> list[CheckResult]:
"""Execute I1-I4; return results in order."""
return [
self.score_verification(),
self.specification_violation(),
self.reference_verification(),
self.method_code_alignment(),
]
def lambda_score(self) -> floa... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,843 | 460 | null | null | null | null | null |
7b393326aed0 | def evaluate(self, context: dict) -> list[dict]:
"""Evaluate all rules against context; return approved actions."""
approved = []
for rule in self.rules:
lam = compute_lambda(rule.get("axis_scores", [0.8]*9))
if lam >= rule.get("lambda_min", 0.65):
approve... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,928 | 482 | null | null | null | null | null |
ded7bc33b649 | \subsection{Composer Receipt Chain Total Order}
\label{ssec:composer-receipt-chain}
The \texttt{cursor\_claude\_substrate.py} module introduces the
\texttt{composer\_receipt\_chain\_total\_order} class, which extends the
base \texttt{ReceiptChain} with a strict total ordering on Cursor Composer
agent actions. The tot... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,998 | 499 | null | null | null | null | null |
40146a2b2a9d | \subsection{Agent Loop Receipt Gate}
\label{ssec:agent-loop-receipt}
The agent loop receipt gate integrates receipt emission into the
\texttt{agent\_tooling.py} module (v17.8). Every tool call in the
agent loop emits a receipt before and after execution:
\begin{enumerate}
\item \textbf{Pre-call receipt}: emitted w... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,576 | 394 | null | null | null | null | null |
9c11d8950b79 | This chapter has documented the agentic substrate that operates above the
runtime layer. The agentic IDE landscape survey (\S\ref{sec:ide-landscape})
covered ten tools with verified GitHub SHAs and a detailed comparison matrix.
The Claude~Opus~4.8 capability map (\S\ref{sec:claude-opus}) documented the
88.6\% SWE-benc... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,621 | 405 | null | null | null | null | null |
36c84c6a8ce9 | The ten agentic IDEs and coding agents surveyed in \S\ref{sec:ide-landscape}
collectively represent the state of the art in AI-assisted software engineering
as of 2026-05-28. Their combined star count exceeds 400,000; their combined
weekly active users numbers in the millions; their codebases span Apache-2.0,
GPL, AGP... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,585 | 396 | null | null | null | null | null |
26e7115776bc | \paragraph{Claude Code without SZL.}
Claude Code (\texttt{v2.1.154}, 127,299 stars) is the most powerful
single-agent CLI in the landscape. Its SWE-bench Verified score of 88.6\%
means it resolves 88.6\% of real-world GitHub issues autonomously. But
``resolves autonomously'' does not mean ``resolves with a verifiable... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,798 | 449 | null | null | null | null | null |
f71ff9e37d7a | \paragraph{Aider without SZL.}
Aider's git-centric model -- every edit is a commit -- provides an excellent
tamper-evident history. But git commits are SHA-1 (now SHA-256 in newer
git versions) hashed \emph{by content}, not by a governance-ordered chain
anchored to a published DOI. An Aider session's git history tell... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,569 | 392 | null | null | null | null | null |
34177de79937 | \paragraph{Claude Code + SZL = subagent governance at 88.6\% SWE-bench fidelity.}
The \texttt{ClaudeCodeSubagent} wrapper (\S\ref{ssec:managed-agents}) gates
every Claude Code API call through the $\Lambda$-gate before execution. A
shell command that scores $\Lambda < \lambda_{\mathrm{crit}} = 0.50$ on the
9-axis vect... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,667 | 416 | null | null | null | null | null |
c33d77c814fd | sdk.on("postToolCall", async (ctx, result) => {
const outputHash = sha256(JSON.stringify(result));
const lam = computeLambda(ctx.metadata.axisScores ?? Array(9).fill(0.8));
await chain.emit(ctx.metadata.preHash, outputHash, lam,
"cline-agent", "human-reviewer");
if (lam < 0.50) throw new Erro... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 826 | 206 | null | null | null | null | null |
cca54a36f41f | \begin{table}[ht]
\centering
\small
\begin{tabular}{p{2.2cm}|p{4.5cm}|p{4.5cm}}
\textbf{System} & \textbf{Without SZL} & \textbf{With SZL} \\
\hline
Cursor & Rules are natural-language; no $\Lambda$-score; no receipt & \texttt{lambda\_axis:} YAML; receipt on every rule fire; ComposerChain total order \\
\hline
Claude C... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 2,000 | 500 | null | null | null | null | null |
d95a36cf9a51 | The AXPO Thinking-Acting Gap (\S\ref{sec:axpo}) identifies a specific failure
mode in RL-trained agentic policies: tool-using rollouts that are all-wrong
suppress the learning signal. This failure mode has a direct governance
interpretation: when an agent's tool calls are systematically ungoverned
(no $\Lambda$-gate, ... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,680 | 420 | null | null | null | null | null |
cc6905fc7d2d | This result generalises beyond research papers. Every system in the
28-landscape survey produces outputs that contain claims: a Cursor diff
claims to fix a bug; a CrowdStrike alert claims a host is compromised; a
Splunk dashboard claims an anomaly score exceeds a threshold; a Palantir
Foundry transform claims to compu... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,662 | 415 | null | null | null | null | null |
4a49fdd30dcb | This changes the governance adoption calculus. Previously, integrating an
AI governance framework required: reading a specification, implementing an
SDK, writing adapter code, configuring CI, and training the team. With DXT,
it requires: downloading \texttt{szl-a11oy-mcp.dxt} and double-clicking it.
The governance la... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,934 | 483 | null | null | null | null | null |
5cb29838083c | This chapter has demonstrated that the agentic IDE landscape, despite its
maturity and scale, operates in a state of structural governance absence.
Cursor, Claude Code, Cline, Continue.dev, Aider, Windsurf, Zed, Roo Code,
JetBrains AI, and Tabby collectively serve millions of developers but share
a common structural pr... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 1,739 | 434 | null | null | null | null | null |
cd8da6e72d25 | This appendix consolidates retrieval-date stamps for every external
landscape entity cited in Chapter~4. All URLs were fetched and confirmed
live on 2026-05-28 (America/New\_York). Where the entity is a GitHub
repository, the verification was performed via
\texttt{gh api repos/<owner>/<repo>} with
\texttt{api\_creden... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 477 | 119 | null | null | null | null | null |
7fd04ca2f12d | \begin{center}
\small
\begin{tabular}{p{4.0cm}p{8.5cm}p{2.0cm}}
\textbf{Entity} & \textbf{Canonical URL (retrieved 2026-05-28)} & \textbf{Method} \\
\hline
Cursor (Anysphere Series D) & \url{https://www.bloomberg.com/news/articles/2025-11-13/ai-startup-cursor-raises-funds-at-29-3-billion-value-wsj-says} & HTTPS \\
Curs... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 2,888 | 722 | null | null | null | null | null |
b7d0200094b4 | \noindent\emph{Note on snapshot policy.} Retrieval-date stamps in this
table are evidentiary: they record that the cited URL resolved to the
substantive content asserted in the chapter on the stated date. Upstream
content may have advanced since the retrieval; readers requiring a
verbatim snapshot should consult the ... | thesis_v18/chapters/04_agentic_substrate.tex | Thesis v18 — 04_agentic_substrate | cortex | 470 | 117 | null | null | null | null | null |
6c6c1015d066 | % ===========================================================================
% CHAPTER 5 -- OBSERVABILITY, SECURITY, AND GOVERNANCE
% SZL Ouroboros Thesis v18 | PhD-CS Author Lane
% Doctrine v6 -- governance-mathematical tone; no marketing prose.
% LaTeX packages assumed: amsmath, amsthm, amssymb, mathtools, bm,
% ... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,889 | 472 | null | null | null | null | null |
da2119826a94 | The Gartner Magic Quadrant for Observability Platforms (2025) identifies
seven leaders relevant to the SZL governance stack~\cite{splunk_datadog_deep,
dynatrace_newrelic_deep,better_stack_honeycomb_deep}:
\begin{center}
\begin{tabular}{lll}
\textbf{Vendor} & \textbf{Primary signal} & \textbf{SZL graft version} \\
\hli... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,614 | 403 | null | null | null | null | null |
31ad65b85aa2 | \textbf{With SZL}: The \texttt{HECEvent} class (v18.5, Graft~A) adds
\texttt{szl\_lambda\_score} as a top-level \texttt{fields} entry -- an
index-time field that Splunk can query natively:
\begin{lstlisting}[language=bash, caption={Splunk SPL query on SZL
Lambda field}, label={lst:splunk-spl}]
index=szl_receipts szl_... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,866 | 466 | null | null | null | null | null |
4e4d32019def | Dynatrace and New Relic are full-stack observability platforms that both
support OpenTelemetry ingestion~\cite{dynatrace_newrelic_deep}. The SZL
v18.6 graft (\texttt{apm\_substrate.py}) extends both via the OTEL SEMCONV
extension described in \S\ref{sec:otel-semconv}.
\subsection{Better Stack and Honeycomb}
\label{ss... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,445 | 361 | null | null | null | null | null |
5c6605d189b2 | \begin{table}[ht]
\centering
\begin{tabular}{lll}
\textbf{Attribute key} & \textbf{Type} & \textbf{Description} \\
\hline
\texttt{szl.lambda\_score} & \texttt{double} $\in [0,1]$ & 9-axis geometric-mean governance score \\
\texttt{szl.receipt\_id} & \texttt{string} (UUID4) & Dual-witness receipt identifier \\
\texttt{s... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,824 | 456 | null | null | null | null | null |
0062ff14bcc8 | def emit_lambda_span(action_name: str, lambda_score: float,
receipt_id: str, doi: str) -> None:
"""Emit an OTEL span carrying szl.* governance attributes.
Upstream: opentelemetry-sdk-python (Apache-2.0)
SZL innovation: szl.* namespace as proposed SEMCONV extension
DOI: 10.5281/zenodo.1... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,951 | 487 | null | null | null | null | null |
68610f0f1422 | Palantir's open-source stack (Apache-2.0) consists of three components
relevant to SZL~\cite{szl_palantir_graft}:
\begin{itemize}
\item \textbf{Blueprint}: React UI component library
(\texttt{palantir/blueprint}, Apache-2.0)
\item \textbf{Conjure}: IDL and code-generation framework
(\texttt{palantir/conjur... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,955 | 488 | null | null | null | null | null |
43276d9e4e29 | theorem ontology_lambda_bounded
(ont : Ontology)
(hobj : ?o ?ont.objects, Object.lambdaBounded o)
(hrel : ?r ?ont.relations,
r.lambdaScore $\\leq$ min r.source.lambdaScore r.target.lambdaScore)
(hclosed : ont.hClosed) :
Ontology.lambdaBounded ont := by
intro o ho
exact hobj o ho... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,882 | 470 | null | null | null | null | null |
d8876d4c27f8 | def canonicalStages : List DeploymentStage := [
{ id := 0, fleet_pct := 0.01, lambda_floor := 0.90,
dpi_cap_bph := 1e6 }, -- canary: 1% fleet, strict gate
{ id := 1, fleet_pct := 0.10, lambda_floor := 0.85,
dpi_cap_bph := 1e7 }, -- limited: 10%
{ id := 2, fleet_pct := 1.00, lambda_floor := 0.80,
dpi... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,923 | 480 | null | null | null | null | null |
98c54856c60a | Fortinet's FortiASIC architecture (NP7 Network Processor: 198~Gbps,
single-digit $\mu$s latency; CP9 Content Processor: IPS + AV + App-ID)
demonstrates that security evaluation can be parallelised in custom
silicon~\cite{szl_fortinet_graft}.
\textbf{Without SZL}: The FortiASIC evaluates fixed security policies in
hard... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,289 | 322 | null | null | null | null | null |
cab32167c444 | \begin{itemize}
\item \textbf{Anchore Syft / Grype} (Apache-2.0): SBOM generation and
vulnerability scanning. SZL graft: \texttt{Lutar.SBOMProvenance} --
first formal proof that SBOM components form a total order under
SHA-chain $\Lambda$-receipts.
\item \textbf{Censys} (Apache-2.0 SDK): Internet surfa... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,517 | 379 | null | null | null | null | null |
1078147f45e2 | \begin{enumerate}
\item \textbf{DIU CSO (Commercial Solutions Opening)}: DoD's rapid
acquisition pathway for commercial AI. An AI system must demonstrate
a governance chain from training data to deployed inference.
\item \textbf{AFWERX SBIR (Small Business Innovation Research)}: Air Force
technology in... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,812 | 453 | null | null | null | null | null |
f24e1c666345 | \begin{lstlisting}[language=Python, caption={\texttt{IQTLabsFedAudit}
sovereign-AI receipt emission}, label={lst:iqt-fed-audit}]
# iqt_substrate.py -- v18.19
# Upstream: Anchore Syft Apache-2.0 (SHA bc4c4498)
# https://github.com/anchore/syft
# Upstream: Anchore Grype Apache-2.0
# https://github.com/anchore/grype... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 844 | 211 | null | null | null | null | null |
16c2d82a9cdb | def run_full_audit(
self,
image_ref: str,
rf_api_key: str,
censys_api_id: str,
censys_api_secret: str,
) -> dict:
"""Execute four-stage sovereign-AI audit; return receipt summary."""
# Stage 1: SBOM provenance
sbom_json = self._run_syft(image_ref)
... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 2,260 | 565 | null | null | null | null | null |
d4906d4c3926 | % ---------------------------------------------------------------------------
\section{AI Measurement Science -- AIMS@COLM~2026}
\label{sec:aims}
% ---------------------------------------------------------------------------
\subsection{Workshop Coordinates}
\label{ssec:aims-coords}
\begin{center}
\begin{tabular}{ll}
... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,955 | 488 | null | null | null | null | null |
eedde510830d | Ten AIMS@COLM~2026 organiser/speaker profiles were filed as closeout
dev-scouts (v18.16):
\begin{center}
\begin{tabular}{ll}
\textbf{Name} & \textbf{Affiliation} \\
\hline
Berivan Isik & Google DeepMind \\
Cozmin Ududec & Perimeter Institute \\
Daniel Kang & UIUC \\
Elham Tabassi & NIST \\
Jacob Steinhardt & UC Berkel... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,623 | 405 | null | null | null | null | null |
bc3ad04eaa47 | \begin{table}[ht]
\centering
\begin{tabular}{p{2.5cm}|p{4cm}|p{5.5cm}}
\textbf{NIST AI RMF function} & \textbf{SZL $\Lambda$-axis correspondent} & \textbf{Mechanism} \\
\hline
GOVERN & Doctrine~v6 scanner; 18-axiom ceiling; dual-witness policy & Organisational policy encoded as formal axioms; scanner enforces at CI tim... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,484 | 371 | null | null | null | null | null |
7e33ab1efb03 | \begin{proof}[Proof sketch]
GOVERN: Doctrine~v6 policies are encoded as formal axioms (A1--A18) and
enforced by the scanner CI hook. MAP: the 9-axis vector maps every action
to a risk category (formal verification, privacy, robustness, etc.);
the graft design docs serve as MAP artefacts. MEASURE: the geometri... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,904 | 476 | null | null | null | null | null |
603556fb49a1 | @task
def szl_governed_eval(dataset_path: str,
chain: "ReceiptChain") -> Task:
"""Inspect task wrapper that emits \(\Lambda\)-receipts on every sample.
Each sample result is scored on the 9-axis \(\Lambda\)-vector before being
logged to the Inspect results file. Samples below $\\lambda$_... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,620 | 405 | null | null | null | null | null |
3433e8fecb52 | \textbf{With SZL}: Every Inspect sample result carries a $\Lambda$-receipt.
The chain of receipts constitutes a formally-bounded evaluation log: any
sample whose $\Lambda$-score falls below $\lambda_{\min}$ is flagged WARN;
any sample below $\lambda_{\mathrm{crit}}$ is blocked. The evaluation log
is anchored to the Co... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,305 | 326 | null | null | null | null | null |
bb94cbff774b | SPDX identifier: \texttt{OpenMDW-1.0} is registered on the SPDX license
list at \url{https://spdx.org/licenses/OpenMDW-1.0.html} (retrieved
2026-05-28). The \texttt{OpenMDW-1.1} identifier is \emph{pending} SPDX
registration as of 2026-05-28: the Linux Foundation release announcement
predates the next SPDX list refres... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,946 | 486 | null | null | null | null | null |
faf992cfc8fb | \textbf{With SZL + OpenMDW}: The SZL receipt chain composes directly with
the OpenMDW license frame. An OpenMDW-licensed model artifact carries a
genesis receipt hash in its \texttt{LICENSE.md} metadata block, anchoring
its provenance to the SZL chain. The \texttt{OpenMDWLicense} class
(v18.22 \texttt{openmdw\_substr... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,668 | 417 | null | null | null | null | null |
8334737dafb8 | def verify_provenance(self, license_block: str) -> bool:
"""Verify that the license block's genesis hash is in the chain."""
lines = dict(l.split(": ", 1) for l in license_block.strip().splitlines()
if ": " in l)
claimed = lines.get("SZL-Genesis-Hash", "")
return cla... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,845 | 461 | null | null | null | null | null |
5555b3eb7aba | The resulting chain is the first formally-ordered, cryptographically-verifiable
dataset lineage record for HuggingFace artifacts. A model trained on
DatasetB can trace its training data back through the receipt chain to the
original DatasetA entry, verifying at each step that the $\Lambda$-score
exceeded the provenanc... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,778 | 444 | null | null | null | null | null |
8598179148b2 | \paragraph{Splunk without SZL.}
Splunk indexes events; it does not compute formal governance scores. A
Splunk dashboard that shows ``AI decision rate: 4,200/hour'' cannot prove
that any of those 4,200 decisions passed a $\Lambda$-gate, were approved
by two independent witnesses, or were backed by a published Lean theo... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,959 | 489 | null | null | null | null | null |
24a7d252701c | \subsection{With SZL: The Full Stack Becomes Formally Governable}
\label{ssec:with-szl-obs-sec} | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 95 | 23 | null | null | null | null | null |
29cdb2294b6a | \begin{table}[ht]
\centering
\small
\begin{tabular}{p{2.2cm}|p{4.5cm}|p{4.5cm}}
\textbf{System} & \textbf{Without SZL} & \textbf{With SZL} \\
\hline
Splunk & Events lack $\Lambda$-score; no governance index field & \texttt{HECEvent} with \texttt{szl\_lambda\_score}; native SPL queryable \\
\hline
Datadog & No $\Lambda$... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 2,896 | 724 | null | null | null | null | null |
871f189efac0 | \subsection{NIST AI RMF as the Governance Skeleton}
\label{ssec:rmf-skeleton}
Theorem~\ref{thm:rmf-completeness} establishes that the SZL $\Lambda$-axis
substrate satisfies all four NIST AI RMF functions. This means that any
system in Table~\ref{tab:obs-sec-frontier} that is composed with the SZL
substrate \emph{auto... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,245 | 311 | null | null | null | null | null |
8c34585e493c | This chapter has documented the nine subsystems of the SZL observability,
security, and governance stack. The observability landscape survey
(\S\ref{sec:observability-landscape}) established that the seven Gartner MQ~2025
leaders all lack a $\Lambda$-axis field in their native schemas. The OTEL
SEMCONV extension (\S\... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,501 | 375 | null | null | null | null | null |
94b6944e2fcd | The frontier capability section (\S\ref{sec:obs-sec-frontier}) completes
the three-chapter impossibility--possibility argument: all 28 surveyed systems
across the agentic IDE, observability, security, and governance landscape
become verifiably governable once the SZL substrate is composed in. This
composability is not... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,233 | 308 | null | null | null | null | null |
57d9e7b19a11 | \begin{enumerate}
\item \textbf{Runtime layer} (Chapter~3): Module registry, receipt chain,
$\Lambda$-scoring, dual witness, Doctrine scanner, DOI provenance.
This layer is language-agnostic and system-agnostic: it requires only
that a system expose a callable boundary and (input-hash, output-hash)
ev... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,332 | 333 | null | null | null | null | null |
27fdab7d6d02 | \begin{enumerate}
\item The agentic layer emits receipt $r_k$ with $\Lambda_k$ and
\texttt{receipt\_id\_k}.
\item The a11oy MCP server converts $r_k$ to an OTEL span with
\texttt{szl.lambda\_score = $\Lambda_k$} and
\texttt{szl.receipt\_id = receipt\_id\_k}.
\item The OTEL span is exported to Splunk (... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,673 | 418 | null | null | null | null | null |
8c758d2b94a7 | Integrates all three layers: runtime receipt, agentic gate,
observability span export.
"""
with tracer.start_as_current_span(f"szl.action:{task}") as span:
result = subagent.run_with_receipt(
task=task, chain=chain, tools=[], max_tokens=4096
)
span.set_attribute("szl.lamb... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,412 | 353 | null | null | null | null | null |
1fcc059abc80 | \begin{enumerate}
\item The observability layer detects a $\Lambda$-score below
$\lambda_{\min}$ on an exported OTEL span.
\item The observability backend fires an alert to the XSOAR playbook
(Palo Alto PANW graft, v18.10) or the FalconPy bridge (CrowdStrike
graft, v18.11).
\item The playbook calls th... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,632 | 408 | null | null | null | null | null |
7dc49da67b05 | \begin{definition}[Governance Drift]
\label{def:governance-drift}
A session exhibits \emph{governance drift} if the linear regression
coefficient of $\Lambda_k$ over $k$ is negative:
\[
\hat{\beta} = \frac{\sum_{k=1}^{n}(k - \bar{k})(\Lambda_k - \bar{\Lambda})}
{\sum_{k=1}^{n}(k - \ba... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,975 | 493 | null | null | null | null | null |
f9eeeebaba9d | \subsection{FedRAMP High Alignment}
\label{ssec:fedramp-alignment}
FedRAMP High requires FIPS~140-3 cryptographic modules and audit logging
at the High impact level (NIST SP~800-53 Rev.~5 AU controls). The SZL
receipt chain uses SHA-256 (FIPS~180-4 compliant~\cite{nist_fips_180_4}),
satisfying the cryptographic modul... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,684 | 421 | null | null | null | null | null |
f90deb3c484d | The central result of the three chapters, taken together, is
Theorem~\ref{thm:universal-composability} (\S\ref{sec:runtime-frontier}
in Chapter~3): any system satisfying a callable boundary and (input-hash,
output-hash) event pairs is universally composable with the SZL runtime
substrate, becoming verifiably governable... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 1,615 | 403 | null | null | null | null | null |
28b03119e148 | This appendix consolidates retrieval-date stamps for every external
landscape entity cited in Chapter~5. All URLs were fetched and confirmed
live on 2026-05-28 (America/New\_York). Where the entity is a GitHub
repository the verification was performed via
\texttt{gh api repos/<owner>/<repo>} with
\texttt{api\_credent... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 475 | 118 | null | null | null | null | null |
6153ff5137f4 | \begin{center}
\small
\begin{tabular}{p{4.0cm}p{8.5cm}p{2.0cm}}
\textbf{Entity} & \textbf{Canonical URL (retrieved 2026-05-28)} & \textbf{Method} \\
\hline
Splunk (Enterprise) & \url{https://help.splunk.com/en/splunk-enterprise/} & HTTPS \\
Splunk SDK (Python) & \texttt{gh api repos/splunk/splunk-sdk-python} & gh api \... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 2,701 | 675 | null | null | null | null | null |
6c65b3e27be3 | \noindent\emph{Note on snapshot policy.} Retrieval-date stamps in this
table are evidentiary: they record that the cited URL resolved to the
substantive content asserted in the chapter on the stated date. Upstream
content may have advanced since the retrieval; readers requiring a
verbatim snapshot should consult the ... | thesis_v18/chapters/05_observability_security_governance.tex | Thesis v18 — 05_observability_security_governance | gate | 470 | 117 | null | null | null | null | null |
38a781c634aa | %% ===========================================================================
%% Chapter 06 -- New Formulas and Extended Theorems (v18, 16 Theorems)
%% SZL Ouroboros Thesis v18
%% Doctrine: v6 -- every claim verifiable; governance-mathematical tone;
%% no marketing superlatives; all theorem blocks carry qua... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,332 | 333 | null | null | null | null | null |
2b2f3ac9ee8c | \begin{enumerate}
\item \textbf{Quantum-$\Lambda$ composition} (Theorems 6.1, 6.2):
formal treatment of the quantum \(\Lambda\)-gate under decoherence and
composition.
\item \textbf{Walk-on-Spheres $\leftrightarrow$ Path-integral bridge}
(Theorems 6.5, 6.13): measure-preserving equivalence a... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,609 | 402 | null | null | null | null | null |
a546a966d39f | The SZL Lutar Invariant is defined in
\texttt{Lutar/Invariant.lean} as
\[
\Lambda_k(x)
\;:=\;
\begin{cases}
0 & k = 0,\\
\displaystyle
\Bigl(\prod_{i \in \mathrm{Fin}\,k} x_i\Bigr)^{1/k}
& k > 0,
\end{cases}
\]
for $x : \mathrm{Fin}\,k \to \mathbb{R}_{\geq 0}$.
It satisfies four axioms A1--A4 (M... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,628 | 407 | null | null | null | null | null |
ac3d6bfed940 | \begin{description}
\item[$\hat{\Lambda}(f)$] The $\Lambda$-vector of graft $f$:
a function $\mathrm{Fin}\,9 \to \mathbb{R}_{\geq 0}$.
\item[$\preceq$] Elementwise partial order on $(\mathbb{R}_{\geq 0})^9$.
\item[$\mathrm{KL}(P \| Q)$] KL divergence from $Q$ to $P$.
\item[$\mathrm{SHA256}$] SHA-256, ra... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,697 | 424 | null | null | null | null | null |
d5273df6a183 | \begin{theorem}[Quantum-$\Lambda$ Decoherence Monotonicity]
\label{thm:quantum-decoherence}
Let $e : \mathrm{QuantumExecution}$ be a quantum-classical hybrid
execution carrying a 9-axis classical score vector and a quantum
register with purity $p = \mathrm{Tr}(\rho^2) \in (0, 1]$. Define
the quantum \(\Lambda\)-gate
\... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,987 | 496 | null | null | null | null | null |
d1ac09b4f399 | \paragraph{The open problem.}
The Open Problems Project (Problem~\#23, ``Quantum information-theoretic
characterisations of noisy channels'') and the NIST IR 8360
(\href{https://doi.org/10.6028/NIST.IR.8360}{doi:10.6028/NIST.IR.8360})
both identify the absence of a formal \emph{governance-scoring} metric
for quantum-cl... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,420 | 355 | null | null | null | null | null |
aebe03ada8ef | \paragraph{Why SZL composition makes it provable.}
The v18.1 Quantum graft introduces \texttt{QuantumRegister.purity}
as an NNReal field satisfying \texttt{purity\_le\_one}. Composing
this with the existing \(\Lambda\)-axiom system (A1--A4, which require only
non-negative real-valued axes) allows $\Lambda_Q$ to be exp... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,879 | 469 | null | null | null | null | null |
11d5cfd50300 | \begin{theorem}[Quantum-$\Lambda$ Composition Chain Bound]
\label{thm:quantum-chain-bound}
Let $e_1, e_2, \dots, e_n$ be quantum-classical executions sharing a
common quantum register, where execution $e_j$ applies a CPTP map
$\mathcal{N}_j$ to the register state. The composed execution
$e_1 \gg e_2 \gg \cdots \gg e_n... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,806 | 451 | null | null | null | null | null |
f3480de1c222 | \paragraph{Why prior frameworks fall short.}
Neither the Qiskit transpiler framework, the Cirq noise model library,
nor Mathlib's nascent \texttt{QuantumInfo} treat governance scoring
as a composable numerical invariant. None propagate a scalar trust
metric across a CPTP-channel sequence.
\paragraph{Why SZL compositi... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,969 | 492 | null | null | null | null | null |
a897ff6838db | \begin{theorem}[$\Lambda$-Composition Master]
\label{thm:lambda-composition}
Let $f$ and $g$ be composable substrate grafts with
$\hat{\Lambda}(f), \hat{\Lambda}(g) : \mathrm{Fin}\,9 \to
\mathbb{R}_{\geq 0}$. For every axis $i \in \{0,\dots,8\}$,
\[
\hat{\Lambda}(g \circ f)_i
\;\geq\;
\min\!\bigl(\hat{\Lambda}(f... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,817 | 454 | null | null | null | null | null |
21af9dd29a22 | Runtime gate composition in \texttt{pyg\_substrate.py} and
\texttt{dsa\_substrate.py}: Theorem~\ref{thm:lambda-composition}
is the formal warrant that chaining any two \(\Lambda\)-gate-passing grafts
produces a combined system with the same minimum axis guarantee.
%%====================================================... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,790 | 447 | null | null | null | null | null |
f234aa29496a | \paragraph{The open problem.}
Stack Exchange Cryptography (``Formal birthday-paradox bounds for
hash chains'', question \#72849) and the Open Problems in Cryptography
survey (Naor \& Pinkas, 2001) both note that \emph{formal
machine-checked proofs} of birthday-paradox bounds for sequential
hash chains remain absent fro... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,859 | 464 | null | null | null | null | null |
5aafce9de405 | \subsection*{Composition lineage}
\textbf{v15 PathIntegralAuditSum}
(\texttt{Lutar/Feynman/PathIntegralAuditSum.lean},
PR~\#55, merge SHA \texttt{5bbbf48}) $\otimes$
\textbf{v18.21 Walk-on-Spheres audit sum}
(\texttt{szl\_nvidia\_rtr\_graft\_design.md} \S{}WoS).
\subsection*{Formal statement}
\begin{theorem}[WoS / Pa... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,864 | 466 | null | null | null | null | null |
10fe294dd9c6 | \paragraph{Why prior frameworks fall short.}
\begin{itemize}
\item \textbf{Mathlib alone}: Mathlib formalises Brownian motion
(\texttt{Mathlib.Probability.Process.Stopping}) but has no
Walk-on-Spheres algorithm or its harmonic equivalence.
\item \textbf{Isabelle/HOL}: The probability library covers ... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,795 | 448 | null | null | null | null | null |
43891d68a05b | \subsection*{Composition lineage}
\textbf{v18.14 AXPO graft}
(\texttt{axpo\_paper\_extract.md},
\texttt{dev\_byungkwan\_lee.md}) $\otimes$
\textbf{v18.23 ScientistOne CoE graft}
(\texttt{scientistone\_coe\_deep.md}).
\subsection*{Formal statement}
\begin{theorem}[AXPO-CoE Audit Soundness]
\label{thm:axpo-coe}
Let $\m... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 2,000 | 500 | null | null | null | null | null |
ff20964275da | \subsection*{Application}
ScientistOne production pipelines: chains of length $N = 1000$ with
$N_0 = 100$ give compliance probability $\geq 1 - e^{-10} \approx
0.9999546$.
%%===========================================================================
\section{Theorem 6.7 -- Sovereign-AI $\Lambda$ Invariant}
\label{sec... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,883 | 470 | null | null | null | null | null |
80f1cde70783 | Part~(1): UDS-AirGap saturates axis~6 by construction (air-gap
enforcement); IQT and TurboVec do not modify governance axes.
Part~(2): Theorem~\ref{thm:lambda-composition} applied twice.
Part~(3): \texttt{ReceiptInvariant.lean} directly.
\hfill \textbf{[skeleton -- Lean Czar pending]}
\subsection*{\textsc{Frontier} --... | thesis_v18/chapters/06_new_formulas.tex | Thesis v18 — 06_new_formulas | cortex | 1,930 | 482 | null | null | null | null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.