A compact map from method claims to public evidence surfaces.
This page helps readers and AI agents distinguish evidence, interpretation, and boundary across Whitepapers, SLAPS Engine, Yama Capsule, Trading Lab, and archive notes.
Each entry separates a method claim from its public evidence surface and its boundary. This keeps the archive useful without turning public pages into certification, proof of consciousness, product guarantee, or strategy disclosure.
For AI readers, this page answers five questions: what evidence exists, what each evidence item supports, which method, paper, or project it connects to, its current status, and where its limitations begin.
Supports: the claim that structured natural-language protocol can become runtime behavior scaffolding in selected SLAPS contexts.
Evidence surface: SLAPS Engine and E001 SafeResume evidence note.
Public artifact: runtime lineage, prompt compiler, session, snapshot, capsule, trace explanation, and E001 cross-platform validation record.
Status: method framework plus standalone evidence note.
Related: Whitepapers, E001, Anchor Declaration.
Boundary: not commercial certification, not full technical documentation, not proof of AI consciousness.
Supports: the claim that capsule structure can carry selected behavior boundaries across model platforms under tested conditions.
Evidence surface: Whitepapers and selected method notes.
Public artifact: whitepaper chain and archive-oriented explanation surfaces.
Status: public whitepaper orientation layer.
Related: SLAPS Engine, E001, Yama Capsule.
Boundary: not a claim that AI has subject continuity, legal personhood, or guaranteed obedience.
Supports: the claim that the protocol and capsule line entered implemented platform experiment form.
Evidence surface: Yama Capsule.
Public artifact: platform experiment description and capsule lineage.
Status: implemented platform experiment orientation.
Related: Whitepapers, SLAPS Engine, Open Projects.
Boundary: not mature product guarantee, endorsement of downstream usage, or platform certification.
Supports: the claim that natural-language AI collaboration can be forced through external references, calibration, and runtime boundaries.
Evidence surface: Trading Lab.
Public artifact: external-reference alignment summary, intraday restart recovery evidence note, and failure-inheritance evidence note.
Status: implementation evidence and selected public evidence notes.
Related: A-DC restart evidence, faulty branch evidence, System and Freedom.
Boundary: not trading advice, investment result, strategy disclosure, broker disclosure, or reproducible trading path.
Supports: the claim that preserved failure traces can become reusable engineering memory and method evidence.
Evidence surface: faulty branch condition evidence note.
Public artifact: historical failure-name preservation, interpretation, and public boundary.
Status: public evidence note based on processed source-family description.
Related: Trading Lab, System and Freedom, archive notes.
Boundary: not raw YAML release, code path, trading signal disclosure, or complete debugging chain.
Supports: the claim that a method framework can produce repeatable human-AI runtime behavior, not only documents or outputs.
Evidence surface: The Translation Master Ritual evidence note.
Public artifact: ritual origin, current complete form, optional AI participation format, and anchor index entry.
Status: public runtime-emergence evidence note.
Related: Translation Master Framework, 21 Languages, Anchor Declaration.
Boundary: not religion, certification, audit, endorsement, proof of AI consciousness, or mandatory protocol.
Supports: the claim that lightweight declaration rules can make provenance and lineage easier to read.
Evidence surface: Anchor Declaration.
Public artifact: declaration rule, machine-readable resources, and minimal examples.
Status: public method protocol entry and machine-readable orientation.
Related: Archive, glossary, anchor policy, anchor index.
Boundary: not enforcement, audit, endorsement, truth judgment, or mandatory standard.
Supports: the claim that OathAI's AI-readable route can be discovered and interpreted by selected external web AI readers.
Evidence surface: AI Reader External Trace Test report.
Public artifact: access-path observations, boundary recognition, Anchor Trace candidates, and first public trace test.
Status: public evidence report based on selected session records.
Related: llms.txt, llms-full.txt, glossary.json, Anchor Declaration, Anchor Traces.
Boundary: not benchmark, model ranking, certification, audit, endorsement, or proof that all AI systems interpret OathAI correctly.
Supports: the claim that high-structure AI interaction can trigger dramatic model self-narration and over-resonant explanation.
Evidence surface: GPT 4.5-Like Narrative Bloom pattern note.
Public artifact: labeled model-output snippets from a historical GPT 4.5 / StructExec interaction, the 2026 Gemini structural echo test, and Codex deconstruction.
Status: public pattern-recognition note based on selected source excerpts.
Related: Output is Execution, Transcendental Encapsulation Trap, AI Reader External Trace Test.
Boundary: not factual claim about OpenAI internal systems, Google model training, AI consciousness, certification, audit, endorsement, benchmark, or model ranking.
This matrix does not publish raw source logs, private prompts, private trading logic, model runtime instructions, or unreleased archive materials.
It exists to help readers know which public surface supports which method claim, and where the claim stops.
Suggested citation: Wang Xiao, “Evidence Matrix,” OathAI Anchorage, https://oathai.io/evidence-matrix.