Patterns, Threats, and Risk Prioritization
Purpose
This document identifies recurring patterns that may emerge when using the JellyLabs foundation and assesses their risk profile.
The goal is not to eliminate all risk, but to:
- make high-impact failures visible early,
- prioritize mitigation effort rationally,
- and prevent accidental canonization of harmful patterns.
This document is explanatory, not normative.
Risk Dimensions
Each pattern is assessed along three axes:
| Dimension | Levels |
|---|---|
| Threat Likelihood | Low, Medium, High |
| Blast Radius | Local (single artifact), Systemic (multiple components), Foundational (authority/trust layer) |
| Damage Severity | Low (confusion), Medium (misuse/rework), High (loss of authority/trust/safety) |
Pattern Risk Tables
Authority & Decision Patterns
| Pattern | Threat | Blast Radius | Damage | Notes |
|---|---|---|---|---|
| Shadow Authority | High | Foundational | High | Most dangerous long-term failure; often invisible |
| Consensus Drift | Medium | Systemic | Medium | Social agreement mistaken for approval |
| Retrospective Justification | Medium | Systemic | High | Governance becomes ceremonial |
AI Collaboration Patterns
| Pattern | Threat | Blast Radius | Damage | Notes |
|---|---|---|---|---|
| Fluent Oracle | High | Foundational | High | Confidence mistaken for correctness |
| Idea Flood | High | Systemic | Medium | Overloads prioritization, not authority |
| Recursive Refinement Trap | Medium | Local | Low | Productivity drag, not governance failure |
Process & Execution Patterns
| Pattern | Threat | Blast Radius | Damage | Notes |
|---|---|---|---|---|
| Conductor Pattern (implicit) | Medium | Systemic | High | Dangerous if canonized silently |
| Checklist Theater | Medium | Local | Medium | Creates false rigor |
| Ledger Drift | Medium | Systemic | Medium | Tasks lose meaning |
Publication & Canon Patterns
| Pattern | Threat | Blast Radius | Damage | Notes |
|---|---|---|---|---|
| Accidental Canonization | High | Foundational | High | Docs override specs socially |
| SEO Gravity | Medium | Systemic | Medium | Explanatory pages out-rank authority |
| Zombie Artifacts | Medium | Systemic | Medium | Conflicting guidance persists |
Social & Temporal Patterns
| Pattern | Threat | Blast Radius | Damage | Notes |
|---|---|---|---|---|
| Founder Context Leak | Medium | Systemic | Medium | Assumptions die with people |
| Institutional Momentum | Low | Systemic | Medium | Inertia replaces judgment |
| Over-Formalization Creep | Medium | Foundational | High | System becomes brittle |
Priority Risk List (Action Order)
Based on likelihood × blast radius × damage, the following patterns deserve the highest attention:
🔴 Tier 1 — Must Be Actively Guarded Against
- Shadow Authority
- Fluent Oracle
- Accidental Canonization
- Over-Formalization Creep
Failure here undermines human authority, the system’s core objective.
🟠 Tier 2 — Monitor and Mitigate
- Retrospective Justification
- Conductor Pattern (implicit)
- Consensus Drift
- SEO Gravity
These erode governance over time if ignored.
🟡 Tier 3 — Acceptable with Awareness
- Idea Flood
- Checklist Theater
- Ledger Drift
- Zombie Artifacts
Manageable through discipline and review.
Mitigation Philosophy
Not every pattern requires:
- a new protocol,
- a new rule,
- or a new checklist.
Many are best handled by:
- explicit naming,
- shared vocabulary,
- and clear boundaries.
Formalization is reserved for patterns that threaten authority or trust.
Relationship to Safety
This document feeds directly into:
- FTL-012 (Safety boundaries and failure containment)
- FTL-011 (AI planning & ideation boundaries)
- FTL-010 (Conductor Pattern evaluation)
It exists to prevent process-level harm, not to address model behavior or ethics broadly.
Authority & Material Safety
Safety in this system depends on a strict separation between public canonical authority and private deliberation.
JellyLabs enforces this separation through Authoritative Material Classification (AMC):
- Public canonical materials are the only source of binding authority.
- Private authoritative materials may justify decisions internally but never impose obligations.
- Exploratory materials carry no authority and are excluded from governance weight.
Any safety failure that allows private or exploratory materials to directly influence canon is considered a foundational breach and must be addressed immediately.
Reader Guidance
If you notice one of these patterns emerging:
- Pause before adding rules
- Identify whether authority is at risk
- Prefer visibility over suppression
- Escalate via proposal if permanence is implied
Closing Note
Patterns are not bugs. Unexamined patterns are.
By documenting them explicitly, the system gains:
- foresight without paranoia,
- rigor without rigidity,
- and safety without theater.