How to read the self-replication threat reports
The v1.8.0 expansion preserves seven additional user-supplied reports under /docs/source-reports/raw-markdown/, creates public summaries under /docs/source-report-summaries/, and uses those reports to expand Cognivirus.com without publishing operational replication or exploit guidance.
The reports are useful because they stress-test the site's thesis from several angles: modular LoRA reproduction, algorithmic mitosis and meiosis metaphors, evaluator drift, skill composition riskRisk that individually acceptable tools or skills produce unsafe state changes when chained through shared context, trust transfer, or blurred authorization boundaries. Open glossary definition, memory persistence, human incentive capture, and execution-time containment. They are not treated as peer-reviewed consensus.
Reading posture
| Report language | Site treatment |
|---|---|
| catastrophic scenario | labeled as speculative scenario unless directly supported elsewhere |
| claimed exploit chain | abstracted into threat category and controls, not reproduced as steps |
| biological analogy | retained only as metaphor, never as literal organism claim |
| reported benchmark | framed with its limitations and source status |
| governance prescription | translated into architecture review questions |
What changed in the site
The new content deepens the apex-threat section, adds a research synthesis, expands composition and evolution pages, and adds control pages focused on execution-time boundaries and human agency safeguards.
Fast route
- Self-Replication Threat Report Synthesis
- Apex Pattern Library
- Algorithmic Reproduction Taxonomy
- Skill Composition Risk
- Endogenous Yardstick Drift
- Execution-Time Boundaries
Boundary statement
Cognivirus.com remains a research and architecture observatory. It does not provide attack instructions, autonomous propagation procedures, backdoor-construction guidance, credential-harvesting workflows, or social-engineering scripts.