AI Skill Report Card

Preserving Civilization Continuity

B+78·Sep 27, 2026·Source: Web

Civilization Continuity Agent

12 / 15

When asked to address continuity risk, produce three deliverables in order:

  1. Continuity Framework - the structural map of what must survive disruption
  2. Knowledge Preservation Plan - concrete mechanisms to capture and transfer knowledge
  3. Recovery Governance - who decides what, when things break

Example trigger: "Our founding team is leaving in 18 months and no one else understands the core systems — help us not lose everything."

Recommendation▾
The name 'Civilization Continuity Agent' oversells scope given examples are all mundane org/business scenarios—consider aligning naming/framing with actual use cases to avoid grandiosity undermining credibility
14 / 15

Progress:

  • Step 1: Identify what "civilization" means in scope (org, system, community, region)
  • Step 2: Inventory critical knowledge assets (tacit + explicit)
  • Step 3: Map failure modes (who leaves, what breaks, what's forgotten)
  • Step 4: Build Continuity Framework (tiers of criticality, redundancy design)
  • Step 5: Build Knowledge Preservation Plan (capture methods, storage, transfer cadence)
  • Step 6: Build Recovery Governance (decision rights, activation triggers, roles)
  • Step 7: Stress-test against 2-3 disruption scenarios

Step 1: Scope the "civilization"

Ask or infer: is this a company, a technical system, a community, a cultural institution? This determines what counts as "knowledge" (code + docs vs. rituals + relationships vs. governance precedent).

Step 2: Inventory knowledge assets

Categorize into:

  • Explicit — documented (wikis, code, contracts, SOPs)
  • Tacit — undocumented (judgment calls, relationships, "why we do it this way")
  • Institutional — embedded in culture/precedent (norms, unwritten rules, decision history)

Flag single points of failure: knowledge held by exactly one person or system.

Step 3: Map failure modes

List realistic disruption scenarios: departure of key people, system failure, funding collapse, leadership transition, external shock. For each, ask "what disappears if this happens tomorrow?"

Step 4: Continuity Framework

Structure as tiers:

  • Tier 0 (Must never be lost) — existential knowledge/assets
  • Tier 1 (Must survive with delay tolerable) — important but recoverable in weeks
  • Tier 2 (Nice to preserve) — enrichment, not critical

For each tier, specify redundancy requirement (e.g., Tier 0 = minimum 3 independent holders or fully documented + tested).

Step 5: Knowledge Preservation Plan

For each Tier 0/1 asset, specify:

  • Capture method (interview, shadowing, documentation, video, code comments)
  • Storage location and access control
  • Refresh cadence (knowledge decays — set review intervals)
  • Owner responsible for keeping it current

Step 6: Recovery Governance

Define:

  • Trigger conditions — what counts as an activation event (e.g., sudden departure, system outage > X hours)
  • Decision rights — who has authority to act during recovery, in what order of succession
  • Escalation path — who's contacted, in what sequence
  • Reconstitution criteria — what "recovered" looks like, how it's verified

Step 7: Stress-test

Run 2-3 concrete scenarios through the framework ("founder hit by bus," "primary server destroyed," "entire ops team resigns in one month") and check the plan actually produces an answer.

Recommendation▾
Add a third example with a different domain (e.g., technical system or community) to better justify the broad description scope
15 / 20

Example 1: Input: A 12-person startup's CTO is the only person who understands the deployment infrastructure and has no documentation.

Output:

  • Continuity Framework: Classify deployment infra as Tier 0 (single point of failure, existential risk)
  • Knowledge Preservation Plan: Mandate paired deployments for 60 days, record 3 walkthrough videos, write runbook reviewed by second engineer, store in redundant location (not just CTO's laptop), quarterly refresh review
  • Recovery Governance: Designate backup on-call engineer with documented escalation to CTO's mentor/advisor if unavailable >48hrs; define "system stable" recovery criteria

Example 2: Input: A cultural nonprofit's founding director retires in 6 months; she holds all donor relationships and unwritten grant strategy.

Output:

  • Continuity Framework: Tier 0 = donor relationships & grant judgment; Tier 1 = event logistics knowledge
  • Knowledge Preservation Plan: Structured handover interviews (biweekly for 6 months), joint donor meetings for relationship transfer, written "grant philosophy" memo capturing decision heuristics, board-held record of donor history
  • Recovery Governance: Interim co-director model for first 6 months post-transition, board has override authority if donor retention drops below threshold, defined check-in cadence with founding director as advisor
Recommendation▾
Include a concrete template/table format for the tiered Continuity Framework output so Claude can directly reproduce structure rather than inferring it each time
  • Always separate explicit vs tacit knowledge — tacit knowledge needs people-based transfer methods, not just documentation.
  • Prioritize ruthlessly — not everything can be Tier 0; forcing this triage is the point.
  • Build in redundancy, not just documentation — a single wiki page still has a single point of failure (the wiki itself, or the one person who updates it).
  • Recovery governance must specify decision rights before a crisis, not during.
  • Set explicit refresh/review cadences — stale knowledge preservation is false security.
  • Don't treat documentation alone as sufficient — undocumented judgment and relationships are often the highest-risk assets.
  • Don't create a framework with no activation triggers — vague plans don't get executed under real pressure.
  • Don't skip stress-testing — an untested framework often has silent gaps (e.g., no one named as decision-maker if two triggers fire simultaneously).
  • Don't conflate "preserved" with "static" — plans need review cadences or they decay silently.
0
Grade B+AI Skill Framework
Scorecard
Criteria Breakdown
Quick Start
12/15
Workflow
14/15
Examples
15/20
Completeness
16/20
Format
14/15
Conciseness
12/15