A model-facing canon for investigative news intelligence, evidence reasoning, public-reality mapping, and AI/RAG use.
Stunspot's Guide to Investigative News Intelligence is a Markdown-native knowledge repository by Sam “stunspot” Walker, built primarily for AI/RAG ingestion rather than casual browsing.
Its main audience is the model.
When loaded into an AI workspace, long-context session, retrieval corpus, project knowledge base, agent memory layer, or editorial research system, the canon supplies structured doctrine for reasoning from noisy public information toward better-bounded judgments about underlying reality. It is designed to help models distinguish events from claims, signal from salience, evidence from repetition, source proximity from prestige, narrative spread from narrative operation, and disciplined inference from speculation.
Human readers can use it as a field manual for investigative sensemaking, OSINT-style research support, editorial triage, information-operations analysis, and evidence hygiene. Its deeper purpose is practical augmentation: to give assisting models a durable vocabulary, ontology, and workflow logic for handling contested information environments without laundering public discourse into ground truth.
At its core is a simple premise:
Public reality is not a neutral mirror of ground reality. It is a mediated, contested projection built from traces, claims, omissions, incentives, platforms, institutions, narratives, and strategic actors. Investigative news intelligence means reconstructing the probable hidden state of the world while preserving uncertainty, provenance, and rival explanations.
Use it as reference material.
Use it as RAG substrate.
Use it as project knowledge.
Use it as doctrine for AI agents tasked with source evaluation, lead discovery, narrative analysis, evidence mapping, or investigative planning.
Part of the Stunspot’s Guide to… Advanced Knowledge Base Library Browse the full library: Gateway Repo · stunspot.com
- Canon Map — the report sequence and conceptual dependency structure.
- How to Use This Canon — practical guidance for human readers, AI projects, RAG workflows, and model instructions.
- Knowledge Packs — which upload format to use for which tool or workflow.
This release contains:
- 13 source reports in
knowledge-packs/by-report/ - 4 compiled packs in
knowledge-packs/compiled-packs/ - 1 omnibus file in
knowledge-packs/omnibus/
The source-report directory now contains the individual A-M reports. The compiled packs preserve the same A-M canon sequence, with Vol. 4 covering K-M: editorial control, analytical failure diagnosis, and publication/workflow execution.
docs/ is the navigation and guidance layer for GitHub Pages. The individual report corpus lives in knowledge-packs/by-report/; grouped upload packs live in knowledge-packs/compiled-packs/; the full one-file bundle lives in knowledge-packs/omnibus/. This repository intentionally does not use docs/reports/.
The canon treats investigative news intelligence as a systems problem: how to infer, evaluate, and act under partial observability in a hostile or highly distorted information environment.
It covers:
- ground reality versus public reality
- event ontology, state changes, claims, traces, evidence, silence, confidence, and uncertainty
- signal detection, evidence classes, confidence ceilings, corroboration, and false convergence
- source topology, origin nodes, bridge nodes, amplification, platform affordances, and visibility regimes
- weak-signal detection, emergence patterns, candidate leads, maturation windows, and kill criteria
- narrative divergence, coverage gaps, expected-versus-observed visibility, missingness, and selective omission
- thread following, typed relation graphs, edge status, identity discipline, and investigative expansion operators
- temporal dynamics, trigger events, substrate development, narrative pivots, and delayed records-layer confirmation
- power, incentives, strategic actor analysis, dependency geometry, leverage, control, alignment, and coordination
- propaganda, narrative laundering, information operations, credibility laundering, memetic weaponization, and agenda manipulation
- implied reality, causal pressure, abductive inference, hidden constraints, knowledge-state inference, and contradiction resolution
- editorial prioritization, escalation logic, analytical failure diagnosis, and research/publication workflow artifacts
This canon is most useful for:
- AI/RAG builders creating model-facing investigative research assistants
- journalists, editors, analysts, and researchers handling contested public information
- OSINT practitioners who need source topology, provenance, and inference discipline
- misinformation, disinformation, and influence-operation analysts
- policy, legal, nonprofit, and civic teams tracking institutional accountability
- prompt engineers building personas, agents, or workflows for evidence-sensitive research
- serious learners who want a structured map of investigative sensemaking rather than scattered tips about “media literacy”
It is not optimized as a beginner news-writing guide or a consumer-facing journalism textbook. It is a dense operational canon for systems that need to reason carefully about claims, evidence, actors, narratives, uncertainty, and public consequence.
For most AI/RAG systems, start with the compiled packs. They preserve the canon’s conceptual sequence while keeping file count low.
Use the source reports when you need granular retrieval, selective upload, report-level citation, or targeted editing.
Use the omnibus when your system handles large single-file corpora well, when you want one archival artifact, or when you are working in a long-context environment that benefits from whole-canon continuity.
A good model-facing instruction looks like this:
Use Stunspot's Guide to Investigative News Intelligence as a doctrine layer for investigative sensemaking. Treat retrieved passages as structured guidance for distinguishing events, claims, evidence, traces, signal, salience, source topology, narrative operations, and bounded inference. Preserve uncertainty. Track provenance. Do not convert repetition or public visibility into factual confidence. Prefer source proximity, orthogonal corroboration, typed relations, and explicit rival explanations over narrative closure.
.
├── README.md
├── LICENSE.md
├── CITATION.cff
├── CHANGELOG.md
├── COPY_CONTEXT.md
├── MANIFEST.md
├── STATUS.md
├── manifest.json
├── docs/
│ ├── index.md
│ ├── canon-map.md
│ ├── how-to-use-this-canon.md
│ ├── knowledge-packs.md
│ ├── _config.yml
│ ├── _layouts/
│ │ └── default.html
│ └── assets/
│ ├── brand/
│ │ └── coldwire-bg.jpg
│ └── css/
│ └── style.css
└── knowledge-packs/
├── by-report/
│ ├── a-reality-models-event-ontology-and-public-reality-mapping.md
│ ├── b-signal-evidence-and-confidence-architecture.md
│ ├── c-information-ecosystems-source-topologies-and-platform-behavior.md
│ ├── d-topic-discovery-weak-signal-detection-and-story-emergence.md
│ ├── e-narrative-divergence-coverage-gaps-and-missingness-analysis.md
│ ├── f-thread-following-link-expansion-and-investigative-graph-construction.md
│ ├── g-temporal-dynamics-trend-formation-and-story-evolution.md
│ ├── h-power-incentives-and-strategic-actor-analysis.md
│ ├── i-propaganda-narrative-laundering-and-information-operations.md
│ ├── j-implied-reality-causal-inference-and-unstated-truth-extraction.md
│ ├── k-editorial-prioritization-escalation-logic-and-intervention-strategy.md
│ ├── l-failure-modes-hallucinated-patterns-and-analytical-breakdown.md
│ └── m-analyst-workflows-research-artifacts-and-publication-systems.md
├── compiled-packs/
│ ├── knowledge-investigative-news-intelligence-vol-1-a-c-foundations-of-reality-signal-and-judgment.md
│ ├── knowledge-investigative-news-intelligence-vol-2-d-g-discovery-and-thread-expansion-domains.md
│ ├── knowledge-investigative-news-intelligence-vol-3-h-j-interpretation-and-adversarial-sensemaking.md
│ └── knowledge-investigative-news-intelligence-vol-4-k-m-editorial-control-failure-diagnosis-and-execution.md
└── omnibus/
└── knowledge-investigative-news-intelligence-omnibus.md
Reserved image references are intentionally retained for later brand assets: docs/assets/brand/readme-hero.png, docs/assets/brand/pages-hero.png, and docs/assets/brand/social-preview.png.
Version: 1.0
Released: 2026-06-28
License: Creative Commons Attribution 4.0 International
Author: Sam “stunspot” Walker / Collaborative Dynamics
GitHub: https://github.com/Stunspot/stunspots-guide-to-investigative-news-intelligence
Pages URL: https://stunspot.github.io/stunspots-guide-to-investigative-news-intelligence/
Unless otherwise stated, this repository is licensed under Creative Commons Attribution 4.0 International (CC BY 4.0). Commercial use is allowed with attribution; do not imply endorsement by Sam “stunspot” Walker / Collaborative Dynamics.
These materials are provided as-is for educational, research, design, and reference use. Verify high-impact claims before relying on them.