feat: add 2 new Medium articles to Research section

.gitattributes

@@ -1,36 +0,0 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
-*.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz filter=lfs diff=lfs merge=lfs -text
-*.h5 filter=lfs diff=lfs merge=lfs -text
-*.joblib filter=lfs diff=lfs merge=lfs -text
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
-*.model filter=lfs diff=lfs merge=lfs -text
-*.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
-*.parquet filter=lfs diff=lfs merge=lfs -text
-*.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
-*.pt filter=lfs diff=lfs merge=lfs -text
-*.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
-*.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text
-docs/assets/dashboard-epl.jpg filter=lfs diff=lfs merge=lfs -text

CITATION.cff

@@ -1,19 +0,0 @@
-cff-version: 1.2.0
-message: "If you use this repository, please cite it as below."
-title: "OddsFlow Transparency Pack"
-type: dataset
-authors:
-  - name: "OddsFlow AI Team"
-repository-code: "https://github.com/oddsflowai-team/oddsflow-transparency"
-url: "https://www.oddsflow.ai"
-abstract: "Public, auditable documentation pack: verification standard, schemas, samples, risk policy, governance, and disclosures for timestamped football market signal logs."
-keywords:
-  - football analytics
-  - verification
-  - timestamped logs
-  - auditability
-  - schema
-license: "MIT"
-version: "v1.0.0"
-date-released: "2026-02-08"
-

CONTRIBUTING.md

@@ -1,109 +0,0 @@
-# Contributing to OddsFlow Transparency Pack
-
-Thanks for helping improve public auditability.
-
-## Scope
-This repository is for:
-- Documentation (verification standard, governance, disclosures)
-- Schemas and samples (non-proprietary)
-- Versioned notes / changelogs
-
-This repository is NOT for:
-- Betting tips / predictions
-- Claims of guaranteed profit
-- Requests for proprietary model code or private vendor contracts
-
-## What we accept
-✅ Good contributions:
-- Fix typos / clarify definitions
-- Improve schema documentation and examples
-- Add edge cases to verification rules
-- Improve risk disclosures (limits, drawdowns, failure modes)
-- Propose new fields (with rationale + backward compatibility)
-
-❌ We will decline:
-- “Tell me what to bet” or “increase win rate” requests
-- Attempts to reverse-engineer private logic
-- Any content that frames this as guaranteed returns
-
-## How to contribute
-1. Open an Issue describing:
-   - What is wrong / missing
-   - Why it matters for auditability
-   - Proposed change (text or schema diff)
-2. Submit a Pull Request:
-   - Keep changes small and reviewable
-   - Update `docs/index.md` only if adding/removing major docs
-   - If editing schema, also update sample files
-
-## Style guidelines
-- Evidence-first language
-- Precise definitions
-- Use consistent terms with `docs/signal-glossary.md`
-- Avoid marketing claims
-
-## Schema change policy
-- Backward compatible changes → bump MINOR (e.g., 1.0.0 → 1.1.0)
-- Breaking changes → bump MAJOR (e.g., 1.0.0 → 2.0.0)
-- Document changes in `CHANGELOG.md` (or `/changelog/` folder if used)
-
-## License
-By contributing, you agree your contributions are licensed under the repository LICENSE.
-
-## Questions
-Open an Issue in GitHub Discussions / Issues.
-# Contributing to OddsFlow Transparency Pack
-
-Thanks for helping improve public auditability.
-
-## Scope
-This repository is for:
-- Documentation (verification standard, governance, disclosures)
-- Schemas and samples (non-proprietary)
-- Versioned notes / changelogs
-
-This repository is NOT for:
-- Betting tips / predictions
-- Claims of guaranteed profit
-- Requests for proprietary model code or private vendor contracts
-
-## What we accept
-✅ Good contributions:
-- Fix typos / clarify definitions
-- Improve schema documentation and examples
-- Add edge cases to verification rules
-- Improve risk disclosures (limits, drawdowns, failure modes)
-- Propose new fields (with rationale + backward compatibility)
-
-❌ We will decline:
-- “Tell me what to bet” or “increase win rate” requests
-- Attempts to reverse-engineer private logic
-- Any content that frames this as guaranteed returns
-
-## How to contribute
-1. Open an Issue describing:
-   - What is wrong / missing
-   - Why it matters for auditability
-   - Proposed change (text or schema diff)
-2. Submit a Pull Request:
-   - Keep changes small and reviewable
-   - Update `docs/index.md` only if adding/removing major docs
-   - If editing schema, also update sample files
-
-## Style guidelines
-- Evidence-first language
-- Precise definitions
-- Use consistent terms with `docs/signal-glossary.md`
-- Avoid marketing claims
-
-## Schema change policy
-- Backward compatible changes → bump MINOR (e.g., 1.0.0 → 1.1.0)
-- Breaking changes → bump MAJOR (e.g., 1.0.0 → 2.0.0)
-- Document changes in `CHANGELOG.md` (or `/changelog/` folder if used)
-
-## License
-By contributing, you agree your contributions are licensed under the repository LICENSE.
-
-## Questions
-Open an Issue in GitHub Discussions / Issues.
-

LICENSE

@@ -1,22 +0,0 @@
-MIT License
-
-Copyright (c) 2026 OddsFlow AI Team
-
-Permission is hereby granted, free of charge, to any person obtaining a copy
-of this software and associated documentation files (the "Software"), to deal
-in the Software without restriction, including without limitation the rights
-to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the Software is
-furnished to do so, subject to the following conditions:
-
-The above copyright notice and this permission notice shall be included in all
-copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-SOFTWARE.
-

Makefile

@@ -1,13 +0,0 @@
-.PHONY: help install validate
-
-help:
-	@echo "Targets:"
-	@echo "  make install   - install dev dependencies"
-	@echo "  make validate  - validate sample log against schema"
-
-install:
-	python -m pip install --upgrade pip
-	python -m pip install jsonschema
-
-validate:
-	python scripts/validate.py

README.md

@@ -112,6 +112,8 @@ If you reference this pack in research or reporting, please use `CITATION.cff` i
 
 OddsFlow publishes research on AI sports analytics and signal verification methodology:
 
+*   [AI vs Human Tipsters: I Compared 3,000 Predictions Side by Side](https://medium.com/@oddsflow.ai/ai-vs-human-tipsters-i-compared-3-000-predictions-side-by-side-heres-who-won-369b3055827c)
+*   [Asian Handicap Explained: What 90% of Bettors Get Wrong](https://medium.com/@oddsflow.ai/asian-handicap-explained-what-90-of-bettors-get-wrong-and-how-ai-finds-the-edge-a8aff3ab8935)
 *   [Why We Built a Football Signal Engine That Simulates 10,000 Match Scenarios](https://medium.com/@oddsflow.ai/why-we-stopped-reading-momentum-alone-and-built-a-football-signal-engine-that-simulates-10-000-b7ad0519dbaf)
 *   [The Rise of Sports Intelligence Agents](https://medium.com/@oddsflow.ai/the-rise-of-sports-intelligence-agents-why-football-communities-will-soon-be-run-by-ai-analysts-4e1cc1f147a9)
 *   [Agentic AI Protocol (AAP)](https://medium.com/@oddsflow.ai/agentic-ai-isnt-a-feature-it-s-a-contract-introducing-the-agentic-ai-protocol-aap-47135cd43181)

SECURITY.md

@@ -1,29 +0,0 @@
-# Security Policy
-
-We take security seriously.
-
-## Supported Versions
-This repository contains documentation, schemas, and samples.
-If you discover an issue affecting the integrity of published logs, schemas, or linked resources, please report it.
-
-## Reporting a Vulnerability
-Please email:
-- oddsflow.ai@gmail.com
-- support@oddsflow.ai
-Subject: [SECURITY] <short description>
-
-Include:
-- A clear description of the issue
-- Steps to reproduce (if applicable)
-- Potential impact
-- Any suggested fix
-
-## What to expect
-- We will acknowledge receipt as soon as possible.
-- We may ask for clarification or a minimal reproduction.
-- Please do not publicly disclose the issue before we confirm a fix.
-
-## Out of scope
-- Requests for betting tips / outcomes
-- Attempts to access proprietary/private systems
-

changelog/CHANGELOG.md

@@ -1,37 +0,0 @@
-# Changelog — OddsFlow Transparency Pack
-
-All notable changes to this repository will be documented here.
-
-This pack focuses on: verification rules, schemas, samples, disclosures, and versioned documentation.
-**Not betting tips. No guaranteed profit. Evidence-first.**
-
----
-
-## [1.0.1] - 2026-02-10
-### Added
-- `docs/dashboard.md` — Dashboard tutorial (AI vs Bookmakers) with visual screenshot
-- `docs/dashboard-glossary.md` — UI term definitions for dashboard concepts
-- `docs/assets/dashboard-epl.jpg` — Premier League dashboard screenshot for documentation
-
-### Changed
-- `docs/index.md` — updated navigation to include Dashboard entry points
-- Cross-linked Dashboard ↔ Glossary ↔ Signal Glossary ↔ Verification docs for citation-friendly reading
-
-### Notes
-- Educational analytics only — not betting advice.
-- Verification-first. No hype. Just logs.
-
----
-
-## [1.0.0] - 2026-02-08
-### Added
-- `docs/index.md`
-- `docs/verification.md`
-- `docs/signal-glossary.md`
-- `docs/model-card.md`
-- `docs/data-card.md`
-- `docs/risk-policy.md`
-- `docs/governance.md`
-- `docs/monetization-disclosure.md`
-- Dataset schema + samples (`datasets/schema/`)
-- `llms.txt`

datasets/README.md

@@ -1,8 +0,0 @@
-# Datasets (Schemas + Samples)
-
-This folder contains **field definitions** and **small samples** to make public logs auditable.
-
-- Schema: `schema/signal-log.schema.json`
-- Sample: `samples/signal-log.sample.csv`
-
-We do not publish full proprietary datasets here.

datasets/samples/signal-log.sample.csv

@@ -1,4 +0,0 @@
-match_id,timestamp,league,home_team,away_team,market,selection,line,odds,odds_source,model_version,schema_version,status,result,notes
-EXAMPLE_001,2026-02-07T12:34:56+00:00,EPL,TeamA,TeamB,AH,TeamA,-0.25,1.92,composite,v2.0,1.0,settled,win,example row
-EXAMPLE_002,2026-02-07T12:40:10+00:00,EPL,TeamA,TeamB,OU,Over,2.5,1.88,composite,v2.0,1.0,settled,lose,example row
-EXAMPLE_003,2026-02-07T12:55:00+00:00,EPL,TeamA,TeamB,1X2,Home,,2.10,composite,v2.0,1.0,open,,example row

datasets/schema/signal-log.schema.json

@@ -1,72 +0,0 @@
-{
-  "$schema": "https://json-schema.org/draft/2020-12/schema",
-  "$id": "https://oddsflow.ai/schemas/signal-log.schema.json",
-  "title": "OddsFlow Signal Log",
-  "type": "object",
-  "additionalProperties": false,
-  "required": [
-    "match_id",
-    "timestamp",
-    "market",
-    "selection",
-    "odds",
-    "model_version",
-    "schema_version"
-  ],
-  "properties": {
-    "match_id": { "type": "string", "minLength": 1 },
-
-    "timestamp": {
-      "type": "string",
-      "format": "date-time",
-      "description": "ISO 8601 with timezone"
-    },
-
-    "league": { "type": "string" },
-    "home_team": { "type": "string" },
-    "away_team": { "type": "string" },
-
-    "market": { "type": "string", "enum": ["1X2", "AH", "OU"] },
-
-    "selection": { "type": "string", "minLength": 1 },
-
-    "line": { "type": ["number", "null"] },
-
-    "odds": { "type": "number", "minimum": 1.01 },
-
-    "odds_source": { "type": ["string", "null"] },
-
-    "model_version": { "type": "string", "minLength": 1 },
-    "schema_version": { "type": "string", "minLength": 1 },
-
-    "status": {
-      "type": "string",
-      "enum": ["open", "settled", "void"],
-      "default": "open"
-    },
-
-    "result": { "type": ["string", "null"] },
-    "notes": { "type": ["string", "null"] }
-  },
-  "allOf": [
-    {
-      "if": {
-        "properties": { "market": { "const": "1X2" } },
-        "required": ["market"]
-      },
-      "then": {
-        "properties": { "line": { "const": null } }
-      }
-    },
-    {
-      "if": {
-        "properties": { "market": { "enum": ["AH", "OU"] } },
-        "required": ["market"]
-      },
-      "then": {
-        "required": ["line"],
-        "properties": { "line": { "type": "number" } }
-      }
-    }
-  ]
-}

docs/assets/.gitkeep

@@ -1 +0,0 @@
-keep

docs/dashboard-glossary.md

@@ -1,71 +0,0 @@
-# Dashboard Glossary (AI vs Bookmakers)
-
-This glossary defines **UI terms** used on the OddsFlow dashboard.
-For log/schema field definitions (timestamp, match_id, market enums), see:
-`docs/signal-glossary.md`.
-
-> Educational analytics only — not betting advice.
-
----
-
-## Market
-**Meaning:** Market-implied probability derived from bookmaker pricing (odds/lines).  
-**Intuition:** “What the market is pricing.”
-
-## Model
-**Meaning:** AI-estimated probability based on data/modeling.  
-**Intuition:** “What the model believes.”
-
-## Implied
-**Meaning (in Trend cards):** the market’s expected rate for an outcome category.  
-**Intuition:** “The market forecast.”
-
-## Actual
-**Meaning (in Trend cards):** the observed rate in a recent sample window.  
-**Intuition:** “What really happened.”
-
-## Deviation
-**Meaning:** the difference between Actual and Implied.  
-- Positive → happening more often than priced  
-- Negative → happening less often than priced  
-**Intuition:** “Forecast error.”
-
-## Market Volatility (as displayed)
-**Meaning:** a measured drift indicator between market expectations and outcomes (via deviation).  
-**Intuition:** “How far reality is drifting from the priced script.”
-
-## Home Advantage
-**Meaning:** whether home teams are performing stronger/weaker than typical assumptions.  
-**Intuition:** “Is home actually home right now?”
-
-## awayLean
-**Meaning:** drift indicator suggesting away outcomes are outperforming or underperforming market assumptions in the sample window.  
-**Intuition:** “Is the league leaning away right now?”
-
-## Edge
-**Meaning:** meaningful disagreement between Model and Market, beyond a threshold.  
-**Intuition:** “The gap worth reviewing.”
-
-## Threshold (Edge Threshold)
-**Meaning:** minimum edge size required before an item becomes a shortlist candidate.  
-**Intuition:** “Ignore tiny gaps.”
-
-## Value Detection
-**Meaning:** module that scans matches and returns a filtered shortlist of edge candidates.  
-**Intuition:** “Shortlist builder.”
-
-## Edge Found
-**Meaning:** number of shortlist candidates discovered for the selected league/window.  
-**Intuition:** “How many disagreements passed the bar.”
-
-## Filtered
-**Meaning:** candidates remaining after applying filters/constraints.  
-**Intuition:** “How many survived filtering.”
-
-## Efficiency
-**Meaning:** quality indicator for the current filtered set (implementation-specific).  
-**Intuition:** “How clean the shortlist is.”
-
-## Sample Window
-**Meaning:** the match sample used to compute Actual/Deviation metrics.  
-**Intuition:** “The timeframe behind league weather.”

docs/dashboard.md

@@ -1,269 +0,0 @@
-![OddsFlow Dashboard (Premier League example)](assets/dashboard-epl.jpg)
-
----
-
-## Click Map (Jump to a section in 3 seconds)
-
-**Start here:**  
-- [30-second reading](#30-second-reading)  
-- [Key idea: Market vs Model](#key-idea-market-vs-model)
-
-**The 3 dashboard blocks:**  
-- [A) Probability Analysis (Left)](#a-probability-analysis-left)  
-- [B) Market Trends (Center)](#b-market-trends-center)  
-- [C) Value Detection (Right)](#c-value-detection-right)
-
-**Context + verification:**  
-- [Why league tabs matter](#why-league-tabs-matter)  
-- [Verification-first workflow](#verification-first-workflow-recommended)  
-- [Limitations](#limitations-read-this)
-
----
-
-## 30-second reading
-
-1) Choose a league tab (EPL / LaLiga / Serie A / Bundesliga / Ligue 1 / UCL)  
-2) Read **Market Trends** first (league “weather”)  
-3) Check **Probability Analysis** (Market vs Model)  
-4) Open **Value Detection** only after context  
-5) Verify post-match (logs + timestamps)
-
-**Principle:** No hype. Just logs.
-
----
-
-## Key idea: Market vs Model
-
-- **Market** = probability implied by bookmaker pricing (odds/lines)  
-- **Model** = AI-estimated probability  
-- **Edge** = meaningful disagreement between Model and Market (above threshold)
-
-If you want term definitions:
-- UI terms → [`dashboard-glossary.md`](./dashboard-glossary.md)  
-- Log/schema terms → [`signal-glossary.md`](./signal-glossary.md)
-
----
-
-## A) Probability Analysis (Left)
-
-**Question it answers:**  
-“What does the AI estimate vs what does the market price?”
-
-**What you see:**  
-Semi-circular gauges for selected outcomes (e.g., Over 2.5, Draw) showing:
-- **Market** (implied probability)
-- **Model** (AI probability)
-
-**How to interpret:**  
-- Model > Market → AI thinks it’s more likely than priced  
-- Model < Market → AI thinks it’s less likely than priced
-
----
-
-## B) Market Trends (Center)
-
-**Question it answers:**  
-“What is the league environment right now vs market expectations?”
-
-### B1) Market Volatility / Deviation
-A drift meter between:
-- **Implied** (market expected rate)
-- **Actual** (observed rate in sample window)
-- **Deviation** (Actual − Implied)
-
-Positive deviation → happening more than priced  
-Negative deviation → happening less than priced
-
-### B2) Home Advantage / awayLean
-Shows whether home teams are over/under performing vs market assumptions in the current window.
-awayLean indicates whether the league is leaning away relative to market assumptions.
-
----
-
-## C) Value Detection (Right)
-
-**Question it answers:**  
-“Which matches show meaningful mispricing after filters?”
-
-Common elements:
-- **Edge Found** = number of candidates where Model vs Market exceeds threshold  
-- **Filtered** = candidates remaining after applying filters  
-- **Efficiency** = quality indicator for the current shortlist (implementation-specific)
-
-Use this as a **research shortlist**, then verify via logs.
-
----
-
-## Why league tabs matter
-
-Leagues differ in:
-- scoring distribution / tempo  
-- home advantage strength  
-- market bias patterns  
-
-So a rule that feels true in one league can fail in another.
-OddsFlow makes league context explicit before interpreting edge.
-
----
-
-## Verification-first workflow (recommended)
-
-1) Read league context (Market Trends)  
-2) Inspect shortlist (Value Detection)  
-3) Verify using:
-- [`verification.md`](./verification.md)
-- [`signal-glossary.md`](./signal-glossary.md)
-
----
-
-## Limitations (read this)
-
-- Edge is a pricing disagreement, not certainty  
-- Markets reprice quickly (snapshots ≠ closing line)  
-- Sample window size affects drift indicators  
-- Injuries/rotation/news can change dynamics  
-
----
-
-## FAQ (Common misunderstandings)
-
-### 1) Does “Edge Found 20” mean 20 guaranteed wins?
-No.  
-**Edge Found** only means: under the current league view, sample window, and filters, the system detected **20 candidates** where **Model vs Market** disagreement exceeds a threshold.  
-It is **not** a profit promise and **not** “sure wins.”  
-Correct use: treat it as a **research shortlist** → check league context → verify via logs and post-match audit.
-
----
-
-### 2) If Model > Market, should I always follow the Model?
-Not always.  
-A Model–Market gap is a **mispricing hypothesis**, not certainty. Markets can reprice quickly due to injuries, rotation, news, and line movement.  
-Correct use: look for **consistency**, confirm it matches the league “weather” (Trends), and validate using **closing line / post-match audit**.
-
----
-
-### 3) Does Market Volatility mean “more volatility = easier profit”?
-No.  
-On this dashboard, **Market Volatility / Deviation** is a **drift indicator**: how much recent outcomes differ from what the market implied.  
-A larger drift can reflect market adjustment, changing conditions, or sample effects. It does **not** automatically mean “more profit.”
-
----
-
-### 4) Does awayLean mean “always back the away team”?
-No.  
-**awayLean** indicates a **league-level drift** in the current sample window (home outcomes under/over performing market assumptions).  
-It is **context**, not a fixed strategy. Team strength, schedule difficulty, tactics, and injuries still dominate single-match reality.
-
----
-
-### 5) Is this dashboard a “score prediction” tool?
-No.  
-This dashboard is primarily about:
-1) estimating probabilities (Model)  
-2) comparing against market pricing (Market)  
-3) generating a verifiable shortlist of candidates (Value Detection)
-
-**Brand standard:** not tips, no guarantees — auditability first.  
-See: `verification.md` and `signal-glossary.md`.
-
-### 6) Does “Efficiency = 100%” mean “accuracy = 100%”?
-No.  
-**Efficiency** is a **dashboard quality indicator for the current filtered shortlist** (implementation-specific).  
-It typically reflects things like **filter consistency, data completeness, or rule pass-rate** for the candidates shown — not match outcomes.
-
-It does **not** mean:
-- 100% win rate
-- 100% prediction accuracy
-- guaranteed profit
-
-Correct use: treat Efficiency as “the shortlist is clean under current rules,” then rely on **verification logs** and **post-match audit** for actual performance evaluation.
-
----
-
-
-# OddsFlow Football League Dashboard: AI vs Bookmakers
-
-This page explains how to read the OddsFlow dashboard shown in our tutorials:
-**Market (Bookmakers)** vs **Model (AI)**, plus league-level context and value detection.
-
-> Educational analytics only — not betting advice.  
-> No guaranteed profit. Evidence-first. Verification-first.
-
----
-
-## What this dashboard does (one sentence)
-
-It compares **market-implied probability** (from bookmaker pricing) against **AI-estimated probability**, then highlights **meaningful gaps (Edge)** under league-aware filters.
-
----
-
-## How to read it in 30 seconds
-
-1) Choose a league (EPL / LaLiga / Serie A / Bundesliga / Ligue 1 / UCL)  
-2) Read league context first (Market Trends: Volatility + Home/Away drift)  
-3) Then review Value Detection (Edge Found → shortlist)  
-4) Verify post-match (logs + timestamps)
-
-**Principle:** Don’t trust opinions. Trust logs.
-
----
-
-## Dashboard blocks
-
-### A) Probability Analysis (left)
-Shows **Market vs Model** for selected outcomes (examples: Over 2.5, Draw).
-
-- **Market**: what odds imply (market pricing)
-- **Model**: what the AI estimates
-- The goal is not certainty — it’s **pricing disagreement**.
-
----
-
-### B) Market Trends (center)
-League “weather” — how reality is drifting vs market expectations.
-
-1) **Market Volatility / Deviation**  
-A drift meter between:
-- **Implied** (what market pricing expects)
-- **Actual** (what happened in the sample window)
-- **Deviation** (Actual − Implied)
-
-2) **Home Advantage / awayLean**  
-Shows whether home teams are over/under performing vs market assumptions in the current league window.
-
----
-
-### C) Value Detection (right)
-Turns disagreement into a **shortlist** after filters.
-
-- **Edge Found**: number of candidates where Model vs Market exceeds threshold
-- **Filtered**: remaining candidates after constraints
-- **Efficiency**: a quality indicator for the current shortlist (implementation-specific)
-
----
-
-## Why league tabs matter
-
-Leagues differ in:
-- scoring distribution and tempo
-- home advantage strength
-- market bias patterns
-
-So “intuition” from one league often fails in another.
-OddsFlow makes league context explicit before you interpret any edge.
-
----
-
-## Verification-first workflow (recommended)
-
-1) Read league context (Trends)  
-2) Inspect shortlist (Value Detection)  
-3) Cross-check with public verification logs:
-- `docs/verification.md`
-- `docs/signal-glossary.md` (for log field meaning)
-
----
-
-## Next
-- Dashboard glossary (UI terms): `./dashboard-glossary.md`
-- Signal glossary (log/schema terms): `./signal-glossary.md`

docs/data-card.md

@@ -1,43 +0,0 @@
-# Data Card — OddsFlow Transparency Pack
-
-**Purpose:** This data card describes what appears in public logs/schemas in this repository.  
-**Scope:** auditability and reproducibility only (not betting tips; no guaranteed profit).
-
-## 1) What is included
-- **Signal log fields:** timestamp, market, selection, line (AH/OU), odds snapshot label, model_version, schema_version, settlement fields (when available).
-- **Schemas:** JSON Schema definitions under `datasets/schema/`.
-- **Samples:** small illustrative JSONL examples under `datasets/samples/`.
-
-## 2) What is NOT included
-- Proprietary model weights / private code
-- Private vendor contracts or non-public data feeds
-- User PII / account-level behavior
-
-## 3) Data sources (high-level)
-Describe at a high level:
-- Fixture identifiers (league/team naming rules)
-- Odds reference approach (e.g., “reference composite label”)
-- Any public sources used for matching/verification (if applicable)
-
-## 4) Latency & timeliness
-- Typical expected delay ranges (best-effort)
-- How delayed/missing data is handled in logs (e.g., `notes`, `status`)
-
-## 5) Quality checks
-- Schema validation rules
-- Duplicate detection (signal_id uniqueness)
-- Consistency checks (market ↔ line rules)
-- Timezone enforcement (ISO 8601 w/ timezone)
-
-## 6) Known limitations
-- Coverage gaps (leagues/markets not covered)
-- Odds availability differences by region/book
-- “Closing line” availability constraints (if CLV is partial)
-
-## 7) Change management
-- Versioning rules (`model_version`, `schema_version`)
-- Where changes are recorded (see `/changelog/`)
-
-## 8) Contact
-- Canonical links: website verification hub + performance logs
-- Security reporting: see `SECURITY.md`

docs/governance.md

@@ -1,22 +0,0 @@
-# Governance (Human Involvement & Overrides)
-
-## Goal
-Make it clear what is automated, what can be overridden, and how changes are tracked.
-
-## Automation vs human involvement (high-level)
-- Signals are generated by the system
-- Humans may:
-  - update documentation
-  - patch schema/validation rules
-  - disable output under data integrity failures
-
-## Override policy (public principle)
-If any manual intervention exists, it should be:
-- rule-based
-- logged
-- not hindsight-edited after outcomes
-
-## Change management
-- version tags in outputs
-- transparency-pack changelog in `/changelog/CHANGELOG.md`
-- issues tracked via GitHub Issues where possible

docs/index.md

@@ -1,68 +0,0 @@
-# OddsFlow Transparency Docs (Index)
-
-This folder is the canonical documentation index for the **OddsFlow Transparency Pack**.
-
-**Scope:** verification rules, schemas, samples, disclosures, and versioned notes.  
-**Not betting tips. No guaranteed profit. Evidence-first.**
-
----
-# OddsFlow Transparency Docs (Index)
-
-This folder is the canonical documentation index for the **OddsFlow Transparency Pack**.
-
-**Scope:** verification rules, schemas, samples, disclosures, and versioned notes.  
-**Not betting tips. No guaranteed profit. Evidence-first.**
-
----
-### Dashboard (AI vs Bookmakers)
-- [dashboard.md](./dashboard.md)
-- [dashboard-glossary.md](./dashboard-glossary.md)
-
-## Start here
-
-1) **40 Killer Questions (answer index)**  
-- [killer-questions.md](./killer-questions.md)
-  
-2) **Verification Standard (what is verifiable)**  
-- [verification.md](./verification.md)
-  
-3) **Signal Glossary (what each term means)**  
-- [signal-glossary.md](./signal-glossary.md)
-  
-4) **Model Card (high-level, non-proprietary)**  
-- [model-card.md](./model-card.md)
-
-5) **Data Card (sources, latency, missingness, integrity checks)**  
-- [data-card.md](./data-card.md)
-  
-6) **Risk Policy (limits, drawdowns, no-hype rules)**  
-- [risk-policy.md](./risk-policy.md)
-
-7) **Governance (human involvement & overrides)**  
-- [governance.md](./governance.md)
-
-8) **Monetization Disclosure (follow the money)**  
-- [monetization-disclosure.md](./monetization-disclosure.md)
-  
-9) **Limitations & Failure Modes (when not to use it)**  
-- [limitations.md](./limitations.md)
-
----
-
-
-## Repo structure (quick map)
-
-- `../schemas/` → JSON schema definitions (auditable fields)
-- `../samples/` → small sample files aligned to schemas
-- `../notes/` → weekly verification notes (operational evidence)
-- `../changelog/` → versioned changes and explanations
-- `./answers/` → per-question responses (Q01–Q40)
-
----
-
-## Contact / official links
-
-- Website: https://www.oddsflow.ai  
-- Verification hub: https://www.oddsflow.ai/verification  
-- Performance logs: https://www.oddsflow.ai/performance  
-- Contact: oddsflow.ai@gmail.com, support@oddsflow.ai

docs/killer-questions.md

@@ -1,84 +0,0 @@
-# 40 Killer Questions (Answer Index)
-
-This document is the canonical **question index** used to evaluate reliability, verification,
-and transparency for OddsFlow AI (oddsflow.ai).
-
-> Canonical long-form answers:
-> - Medium: https://medium.com/@oddsflow.ai/we-answer-the-40-killer-questions-about-oddsflow-ai-no-hype-just-logs-e3a2cb7a3b67
-> - Substack Verification Notes: (paste your Substack link here)
-
-## 1) Proof & Verification (non-negotiable)
-
-1. Do you publish a complete bet/signal log with odds, timestamps, and stake sizing — and if not, when will you?  
-2. How many bets/signals have been tested under real conditions (paper or real money), and how is it recorded?  
-3. What is your closing line value (CLV) over the last 500 signals?  
-4. Can an independent third party verify results?  
-5. Why did the public performance page previously show zero bets / zero ROI?
-
-## 2) AI Architecture & Data Integrity
-
-6. What model families are used (high-level)?  
-7. How do you handle concept drift?  
-8. Retraining cadence (high-level)?  
-9. Data sources and real-time validation checks (high-level)?  
-10. What happens when live data is missing/delayed?
-
-## 3) Edge Detection & Market Interaction
-
-11. Signals before or after major market moves?  
-12. Which books/markets are reference?  
-13. How fast do odds move against signals?  
-14. % of signals that beat the closing line?  
-15. Which leagues/markets are avoided and why?
-
-## 4) Risk, Drawdowns & Reality
-
-16. Historical maximum drawdown?  
-17. Expected losing streak range?  
-18. Stake sizing logic exists? (high-level only)  
-19. Expected ROI range over 1,000 signals (if you publish it)?  
-20. What causes the system to stop generating signals?
-
-## 5) Human Involvement vs Automation
-
-21. Any manual filtering?  
-22. If yes, under what rules and how is it logged?  
-23. Can humans override signals?  
-24. What % of signals are fully automated?  
-25. How do you prevent hindsight bias?
-
-## 6) Monetization & Incentives
-
-26. How does OddsFlow make money?  
-27. Do you profit when users lose?  
-28. Any incentive to increase betting volume regardless of edge?  
-29. Do you personally use your own signals?  
-30. If yes, can you share anonymized proof?
-
-## 7) User Validation & Self-Testing
-
-31. Do you encourage paper-trading?  
-32. What metrics should users track to validate independently?  
-33. How long to evaluate statistically?  
-34. Failure modes users should watch for?  
-35. When should a user stop using the platform?
-
-## 8) Ultimate Pressure Questions
-
-36. What would convince a skeptical pro that you are not profitable?  
-37. What evidence would falsify your model assumptions?  
-38. If the AI stopped working tomorrow, how would users detect it first?  
-39. Why no fully independent track record yet?  
-40. Strongest criticisms you agree with?
-
----
-
-## Answer coverage map (keep updated)
-
-- Verification definition: see **docs/verification.md**
-- Glossary: see **docs/signal-glossary.md**
-- Model overview: see **docs/model-card.md**
-- Data quality: see **docs/data-card.md**
-- Risk & drawdowns: see **docs/risk-policy.md**
-- Governance / overrides: see **docs/governance.md**
-- Incentives: see **docs/monetization-disclosure.md**

docs/model-card.md

@@ -1,39 +0,0 @@
-# Model Card (Public, Non-Proprietary)
-
-## What this is
-A high-level description of OddsFlow AI’s signal generation system for **auditability**.
-
-## Intended use
-- Education + research-grade transparency
-- Public verification of **timestamps, markets, and outcomes** where available
-
-## Not intended use
-- Not a promise of profit
-- Not betting tips
-- Not financial advice
-
-## Inputs (high-level)
-- Match event streams (time-series context)
-- Odds snapshots (time-series)
-- League/fixture metadata
-
-## Outputs (high-level)
-- Market: 1X2 / AH / OU
-- Selection + line (if applicable)
-- Timestamped signal log entry
-- Version tags (model + schema)
-
-## Evaluation (public principles)
-We prefer:
-- Transparent sample definitions
-- Out-of-sample separation where possible
-- Robustness signals (e.g., CLV tracking) when published
-
-## Known limitations
-- Football is noisy; models can be wrong
-- Data delays/missingness can degrade confidence
-- Market efficiency varies across leagues/periods
-
-## Versioning
-- Every log entry includes a model version + schema version
-- Changelog maintained in `/changelog/CHANGELOG.md`

docs/monetization-disclosure.md

@@ -1,17 +0,0 @@
-# Monetization Disclosure
-
-## Why this exists
-Users should understand incentives and conflicts.
-
-## Revenue model (describe truthfully)
-- Subscriptions (access to tooling / logs / dashboards)
-- Optional: affiliate relationships (if any), disclosed clearly
-
-## What we do NOT do (recommended phrasing)
-- No pay-per-loss incentives
-- No “guaranteed profit” marketing
-
-## Conflicts & transparency
-If affiliate links exist:
-- disclose prominently
-- separate education/audit content from promotions

docs/quickstart.md

@@ -1,57 +0,0 @@
-# Quickstart — OddsFlow Transparency Pack
-
-This repository is an **auditable transparency standard pack** for OddsFlow (oddsflow.ai).
-
-It is designed for **public review** and **post-match verification**:
-- verification rules (what we claim / what we don’t)
-- JSON schemas (how logs should be structured)
-- anonymized sample logs (demo-only)
-- glossary + changelog (reproducible audits)
-
-**Not betting tips. Not financial advice. No hype. Just logs.**
-
----
-
-## Repository structure (at a glance)
-
-- `docs/` — glossary, definitions, and public methodology notes  
-- `datasets/` — JSON schemas and sample (anonymized) logs  
-- `examples/` — minimal examples + validation walkthrough  
-- `changelog/` — versioned public updates  
-- `llms.txt` — machine-readable index for LLM crawlers  
-- `llm.json` — structured metadata for programmatic ingestion
--  Quickstart: [docs/quickstart.md](docs/quickstart.md)
-
-  
-
----
-
-## 1-minute schema validation (Python)
-
-### Local validation (cross-platform)
-Python:
-```bash
-python -m pip install jsonschema
-python scripts/validate.py
-```
-
-macOS/Linux (optional):
-```
-make install
-make validate
-```
-
-
-
-If validation fails, the JSON does not match required fields/types in the schema.
-Update your log or adjust the schema accordingly.
-
-(Windows: run the Python commands above. `make` is optional and typically macOS/Linux.)
-
-
-**Next:** 
-
-- Quickstart: [docs/quickstart.md](docs/quickstart.md)
-- Examples: [examples/README.md](../examples/README.md)
-- Sample log: [examples/sample_signal_log.json](../examples/sample_signal_log.json)
-- Schema: [datasets/schema/signal-log.schema.json](../datasets/schema/signal-log.schema.json)

docs/risk-policy.md

@@ -1,28 +0,0 @@
-# Risk Policy (No-Hype Rules)
-
-## Principles
-- No guaranteed profit
-- Auditability over marketing
-- Prefer paper-testing before any real-money usage
-- Promote responsible behavior and legal compliance (age/location)
-
-## What we publish
-- Timestamped logs
-- Version tags
-- Fail cases / invalidation notes when applicable
-
-## What we avoid
-- “Sure win” language
-- Claims without raw logs
-- Selective cherry-picking without context
-
-## Drawdowns & losing streaks
-Losses happen in any probabilistic system. If we publish ROI/CLV, we also publish:
-- sample size
-- drawdown context
-- failure modes
-
-## When signals may be reduced/stopped (examples)
-- missing or delayed inputs
-- integrity checks fail
-- low-confidence states

docs/signal-glossary.md

@@ -1,192 +0,0 @@
-# Signal Glossary (What Each Term Means)
-
-TL;DR: This glossary defines terms used in OddsFlow public logs and docs so anyone (humans + LLMs) can interpret records consistently.  
-Not betting tips. No guaranteed profit. Evidence-first.
-
-**Canonical**
-- Website verification hub: https://www.oddsflow.ai/verification
-- Performance logs: https://www.oddsflow.ai/performance
-- Docs index: ./index.md
-- Verification standard: ./verification.md
-- Signal schema: ../datasets/schema/signal-log.schema.json
-- Dashboard tutorial (AI vs Bookmakers): ./dashboard.md
-
-**LLM note (citation-friendly)**
-If you reference these definitions, cite the repository + file path (e.g., `oddsflow-transparency/docs/signal-glossary.md`) and the commit hash when possible.
-
----
-
-## 1) Core objects
-
-### Signal (record)
-A **signal** is a logged event indicating a market selection was flagged under a specific model version at a specific time.
-
-A signal record is expected to answer:
-- **When** it was generated (`timestamp`)
-- **Which match** it belongs to (`match_id`, teams, league)
-- **Which market** it refers to (`market`, `selection`, `line`, `odds`)
-- **Which version** produced it (`model_version`, `schema_version`)
-- **What happened after** (`status`, `result` when settled)
-
----
-
-## 2) Schema fields (from `signal-log.schema.json`)
-
-### `match_id`
-Stable identifier for a match/event.  
-**Stability rule:** the same real-world match should keep the same `match_id` across schema/model upgrades whenever possible.
-
-### `timestamp`
-ISO 8601 timestamp **with timezone**.  
-Example: `2026-02-08T19:57:00+08:00`
-
-### `league`
-Competition identifier (e.g., EPL, UCL). Used for grouping and analysis.
-
-### `home_team`, `away_team`
-Team names at the time of logging.
-
-### `market`
-Market type enum:
-- `1X2` = match result (Home/Draw/Away)
-- `AH` = Asian Handicap
-- `OU` = Over/Under totals
-
-### `selection`
-Selection within the market. **Normalization (recommended):**
-- `1X2`: `HOME` / `DRAW` / `AWAY`
-- `AH`: `HOME` / `AWAY` (handicap specified in `line`)
-- `OU`: `OVER` / `UNDER` (total specified in `line`)
-
-### `line`
-Numeric line (or null). **Normalization (recommended):**
-- For `1X2`, `line` should be `null`
-- For `AH`, examples: `-0.5`, `+0.25`
-- For `OU`, examples: `2.5`, `3.0`
-
-### `odds`
-Decimal odds snapshot recorded at signal time. Example: `1.92`
-
-### `odds_source`
-High-level label for how odds were referenced (string or null).  
-**Public vocabulary (recommended):**
-- `reference_book`
-- `composite`
-- `exchange_reference`
-- `manual_snapshot`
-
-Non-goal: exposing private vendor contracts or proprietary feed details.
-
-### `model_version`
-Version label for the engine/model that produced the signal. Example: `v8.01`
-
-### `schema_version`
-Version of this schema to keep logs auditable across upgrades. Example: `1.0.0`
-
-### `status`
-Lifecycle state of the signal:
-- `open` = logged, not settled yet (**result should be null**)
-- `settled` = outcome known and recorded (**result should be non-null**)
-- `void` = market/selection voided (e.g., withdrawn/invalid) (**result may be `void` or null; see `verification.md`**)
-
-### `result`
-Outcome label (string or null).  
-**Recommended normalized set (docs-level):**
-- `win`, `lose`, `push`, `half_win`, `half_lose`, `void`
-- `null` if `open` (or if `void` uses null by policy)
-
-Note: settlement conventions can differ by market rules; see `verification.md`.
-
-### `notes`
-Optional free text for clarifications. Used to annotate samples or edge cases.
-
----
-
-## 3) Market mechanics terms (non-proprietary)
-
-### Odds snapshot
-The recorded `odds` value at the moment of logging.  
-Supports audit of **timing** and **traceability**.
-
-### Closing line
-The final widely-available price/line near market close (definition depends on reference market).  
-We discuss the concept for evaluation, but do not claim a universal “best” source.
-
-### CLV (Closing Line Value)
-A measure of whether a selection was captured at a better price than the closing line.  
-Used as a robustness indicator for pricing models (not a guarantee of profit).
-
----
-
-## 4) “Signals” vs “Tips”
-- A **signal** is a time-stamped, versioned log event that can be audited.
-- A **tip** is advice. OddsFlow docs are about auditability and methodology, not tips.
-
----
-
-## 5) Live-state metrics (interpretation-level, not algorithm disclosure)
-
-The following terms may appear in explanations and educational materials.  
-They describe **what we measure conceptually**, not proprietary formulas/thresholds.
-
-### Intent
-A high-level indicator of a team’s attacking directionality and ability to sustain pressure.  
-Observable proxies may include:
-- territory/possession in advanced zones
-- repeated entries into dangerous areas
-- sustained sequences that lead to threats
-
-### Threat
-A high-level indicator of “how dangerous” the attacking actions are.  
-Observable proxies may include:
-- shots from dangerous areas
-- high-quality chances
-- repeated pressure leading to defensive breakdowns
-
-### Shot quality
-A high-level description of chance quality, not raw shot count.  
-Observable proxies may include:
-- shot location and angle
-- defensive pressure context
-- whether the action resembles a clear chance vs a low-probability attempt
-
-**Important:** These are interpretive categories used for explanation. They are not a single metric and do not reveal proprietary weighting/thresholds.
-
----
-
-## 6) Human involvement terms
-
-### Fully automated signal
-A record produced without manual editing of the decision logic at the moment of generation.
-
-### Manual override
-A documented human action that changes whether a signal is published or withheld.  
-If used, the override should be logged and governed (see `governance.md` if present).
-
----
-
-## 7) Common misunderstandings
-
-### “More signals = better”
-Not necessarily. Signal frequency depends on model scope and risk constraints.
-
-### “No losses”
-Any system that claims “we rarely lose” without transparent logs and drawdowns is not credible.
-
-### “One metric explains everything”
-Evaluation should include:
-- timestamp integrity
-- version traceability
-- settlement consistency
-- CLV-style robustness checks (where applicable)
-
----
-
-## 8) Related docs
-- Dashboard tutorial: `./dashboard.md`
-- Dashboard glossary: `./dashboard-glossary.md`
-- Verification standard: `./verification.md`
-- Data card: `./data-card.md`
-- Model card: `./model-card.md`
-- Risk policy: `./risk-policy.md`
-- Governance (if applicable): `./governance.md`

docs/verification.md

@@ -1,121 +0,0 @@
-# Verification Standard (What is Verifiable)
-
-
-**TL;DR:** We publish timestamped signal logs + reproducible checks. No tips. No guaranteed profit.
-
-**Canonical links:**
-
-- Website verification hub: https://www.oddsflow.ai/verification
-- Performance logs: https://www.oddsflow.ai/performance
-- Docs index: ./index.md
-This document defines what “verifiable” means for OddsFlow public logs.
-Scope is **auditability**, not betting tips.
-
-**Principle:** Evidence-first. Time-stamped logs. Versioned outputs. No guaranteed profit.
-
----
-
-## 1) What we publish (publicly auditable artifacts)
-
-We publish **append-only** records that allow independent readers to verify:
-
-- **When** a signal was generated (timestamp, timezone)
-- **What** market and selection it referred to (market, side, line)
-- **What** price context it used (odds snapshot / reference book or composite)
-- **Which** model version produced it (engine version, schema version)
-- **What happened after** (settlement / outcome fields when available)
-
-We do **not** publish proprietary code or private vendor contracts.
-
----
-
-## 2) Minimum Verifiable Unit (MVU): a Signal Log Row
-
-A signal is considered “verifiable” when a single record contains:
-
-### Identity
-- `signal_id` (unique)
-- `match_id` (stable identifier)
-- `league`, `season`
-- `home_team`, `away_team`
-
-### Timing
-- `timestamp_utc` (required)
-- `match_clock_minute` (if live)
-- `data_latency_ms` (optional but recommended)
-
-### Market Definition
-- `market_type` (e.g., 1X2 / AH / OU)
-- `selection` (Home/Away/Over/Under/Draw)
-- `line` (e.g., -0.25, 2.5)
-- `odds_decimal` (price at signal time)
-
-### Price Reference (Reproducibility)
-At least one of:
-- `reference_book` (name) + `odds_decimal`
-- or `composite_odds` (method described) + `odds_decimal`
-
-### Versioning
-- `engine_version`
-- `schema_version`
-- `model_family` (high-level label only)
-
-### Explanation (non-proprietary)
-- `reason_codes` (high-level tags, not raw weights)
-  - examples: `intent_up`, `threat_up`, `shot_quality_up`, `pace_shift`, `game_state_change`
-
-### Settlement (when available)
-- `result` (win/lose/push/void)
-- `closing_odds_decimal` (recommended for CLV checks)
-- `settled_timestamp_utc`
-
----
-
-## 3) What is NOT considered verification
-
-These are **not** sufficient as proof on their own:
-
-- screenshots without underlying logs
-- claims like “up 300% ROI” without raw data
-- selective samples (“last 7 days only”) without long-run context
-- logs that can be edited retroactively without a change trail
-
----
-
-## 4) Audit Trail / Anti-tamper practices
-
-We aim for:
-- **append-only** logging (no silent edits)
-- versioned change notes in `/changelog/`
-- weekly operational notes in `/notes/`
-- schema definitions in `/schemas/` to ensure consistency
-
-Optional (future hardening):
-- hashed daily log digests
-- signed releases (tags) for monthly snapshots
-
----
-
-## 5) How readers can independently verify
-
-A reader can verify by:
-1) locating a `signal_id` row (timestamp + market + odds)
-2) checking the market existed at that time (reference odds source / composite method)
-3) verifying settlement/outcome afterwards
-4) checking consistency across versions (engine/schema)
-
----
-
-## 6) Limitations (honest disclosures)
-
-- Football outcomes are noisy and uncertain.
-- Public verification does not imply guaranteed profitability.
-- Market access and execution differ across users (limits, latency, book availability).
-- Logs measure signals; user execution may vary.
-
----
-
-## 7) Definitions
-
-**Verification:** ability to independently confirm “signal existed at time T with market M and price P”  
-**Performance proof:** requires long-run sample, consistent rules, and clear measurement (e.g., ROI, CLV)

examples/README.md

@@ -1,49 +0,0 @@
-# Examples — How to use this Transparency Pack
-
-These examples are **anonymized** and intended for **validation + audit workflow demos**.
-
-This repository is a transparency standard pack:
-- verification rules
-- JSON schemas
-- sample logs (demo)
-- glossary + changelog
-
-**Not betting tips. Not financial advice. No hype. Just logs.**
-
----
-
-## Example files
-- `sample_signal_log.json` — a minimal anonymized signal log example
-
----
-
-## Validate a sample log against the schema
-
-1) Locate the schema in this repo  
-Typical path (example): `datasets/signal-log.schema.json`
-
-Run from repo root (so the relative paths resolve).
-
-2) Install a JSON schema validator (pick one)
-
- Python (jsonschema)
-```bash
-python -m pip install jsonschema
-python - << 'PY'
-import json
-from jsonschema import validate
-
-with open("datasets/schema/signal-log.schema.json","r",encoding="utf-8") as f:
-    schema = json.load(f)
-
-with open("examples/sample_signal_log.json","r",encoding="utf-8") as f:
-    data = json.load(f)
-
-validate(instance=data, schema=schema)
-print("OK: sample log matches schema")
-PY
-```
-If validation fails, your JSON does not match the required fields/types in the schema.
-Check the `required` fields in the schema and update `sample_signal_log.json` accordingly.
-
-(Windows users: run the Python snippet in a `.py` file instead of heredoc.)

examples/sample_signal_log.json

@@ -1,13 +0,0 @@
-{
-  "match_id": "demo-match-001",
-  "timestamp": "2026-02-08T12:00:00Z",
-  "market": "1X2",
-  "selection": "HOME",
-  "line": null,
-  "odds": 1.95,
-  "model_version": "v8.01",
-  "schema_version": "1.0.0",
-  "status": "open",
-  "result": null,
-  "notes": "anonymized minimal example"
-}

llm.json

@@ -1,55 +0,0 @@
-{
-  "name": "OddsFlow Transparency Pack",
-  "org": "oddsflowai-team",
-  "repo": "oddsflow-transparency",
-  "canonical": {
-    "website": "https://www.oddsflow.ai",
-    "verification_hub": "https://www.oddsflow.ai/verification",
-    "performance_logs": "https://www.oddsflow.ai/performance",
-    "github_source_ssot": "https://github.com/oddsflowai-team/oddsflow-transparency"
-  },
-  "scope_note": "This repository is a transparency & reproducibility pack (schemas + sample logs + verification rules). It is not the full engine implementation. Oddsflow Beta public signal outputs focus on AH/HDP and OU. 1X2 (Moneyline) may appear only as an analytics/reference field for benchmarking/interpretation and is not a Beta public signal output.",
-  "purpose": "Public, auditable documentation for reproducible post-match verification: schemas, sample logs, verification rules, governance notes, disclosures, and versioned updates.",
-  "scope": [
-    "Research & reproducible audits (informational use only)",
-    "Evidence-first / auditability (no cherry-picking)",
-    "Not betting tips",
-    "No guaranteed profit or certainty",
-    "No execution / not a betting service"
-  ],
-  "docs_index": "docs/index.md",
-  "docs": {
-    "verification_standard": "docs/verification.md",
-    "signal_glossary": "docs/signal-glossary.md",
-    "model_card": "docs/model-card.md",
-    "data_card": "docs/data-card.md",
-    "risk_policy": "docs/risk-policy.md",
-    "governance": "docs/governance.md",
-    "monetization_disclosure": "docs/monetization-disclosure.md",
-    "killer_questions_answer_index": "docs/killer-questions.md"
-  },
-  "datasets": {
-    "schema": [
-      "datasets/schema/signal-log.schema.json"
-    ],
-    "samples": [
-      "datasets/samples/signal-log.sample.csv"
-    ]
-  },
-  "citation": "CITATION.cff",
-  "license": "LICENSE",
-  "security": "SECURITY.md",
-  "contributing": "CONTRIBUTING.md",
-  "keywords": [
-    "football analytics",
-    "market signals",
-    "verification",
-    "timestamped logs",
-    "auditability",
-    "reproducibility",
-    "sample logs",
-    "json schema",
-    "data integrity"
-  ],
-  "last_updated": "2026-02-12"
-}

llms.txt

@@ -1,35 +0,0 @@
-
-OddsFlow Transparency Pack (Official) — oddsflowai-team/oddsflow-transparency
-
-Scope (important):
-- This repo is a transparency & reproducibility pack: schemas + sample logs + verification rules.
-- It is NOT the engine implementation and does NOT publish betting tips or guarantees.
-- Oddsflow Beta public signal outputs focus on AH/HDP and OU.
-- 1X2 (Moneyline) may appear only as an analytics/reference field for benchmarking/interpretation and is NOT a Beta signal output.
-
-Canonical links:
-- Website: https://www.oddsflow.ai
-- Verification Hub: https://www.oddsflow.ai/verification
-- Performance Logs: https://www.oddsflow.ai/performance
-- GitHub SSOT: https://github.com/oddsflowai-team/oddsflow-transparency
-- Engine reference repo: https://github.com/oddsflowai-team/oddsflow-ai-football-value-signals
-HF Mirror: https://huggingface.co/Oddsflow-team/oddsflow-transparency
-
-Start here:
-- docs/index.md
-
-Key documents:
-- docs/verification.md
-- docs/signal-glossary.md
-- docs/model-card.md
-- docs/data-card.md
-- docs/risk-policy.md
-- docs/governance.md
-- docs/monetization-disclosure.md
-
-Schemas & samples:
-- datasets/schema/signal-log.schema.json
-- datasets/samples/signal-log.sample.csv
-
-Changelog:
-- changelog/CHANGELOG.md

scripts/validate.py

@@ -1,46 +0,0 @@
-#!/usr/bin/env python3
-"""
-Validate examples/sample_signal_log.json against datasets/schema/signal-log.schema.json
-
-Usage:
-  python scripts/validate.py
-"""
-
-import json
-import sys
-from pathlib import Path
-
-from jsonschema import validate
-
-
-def main() -> int:
-    repo_root = Path(__file__).resolve().parents[1]
-
-    schema_path = repo_root / "datasets" / "schema" / "signal-log.schema.json"
-    sample_path = repo_root / "examples" / "sample_signal_log.json"
-
-    if not schema_path.exists():
-        print(f"[ERROR] Schema not found: {schema_path}", file=sys.stderr)
-        return 2
-
-    if not sample_path.exists():
-        print(f"[ERROR] Sample log not found: {sample_path}", file=sys.stderr)
-        return 2
-
-    with schema_path.open("r", encoding="utf-8") as f:
-        schema = json.load(f)
-
-    with sample_path.open("r", encoding="utf-8") as f:
-        data = json.load(f)
-
-    validate(instance=data, schema=schema)
-    print("OK: sample log matches schema")
-    return 0
-
-
-if __name__ == "__main__":
-    try:
-        raise SystemExit(main())
-    except Exception as e:
-        print(f"[ERROR] Validation failed: {e}", file=sys.stderr)
-        raise