Spanda

Recovery Coverage

Status: Stable · Horizon: NOW · Priority: P0

Measure recovery readiness — which known failures have recovery plans and which paths are missing.

CLI

spanda recovery-coverage rover.sd
spanda recovery-coverage rover.sd --json
spanda recovery-coverage rover.sd --format markdown

Core types

Type	Purpose
RecoveryCoverageReport	Full report with coverage percentage
RecoveryScenario	Named failure mode with expected recovery
RecoveryGap	Failure without adequate recovery path

Known failure catalog (v1)

Derived from hardware profile, capability matrix, connectivity policies, and fleet configuration:

• GPS loss, battery critical, connectivity loss
• Sensor failure (lidar, IMU, camera)
• Actuator failure, provider timeout
• Fleet peer loss, swarm member failure
• Package/provider unavailable
• Human approval timeout

Output

{
  "coverage_pct": 64,
  "known_failures": 11,
  "covered": 7,
  "partially_covered": 2,
  "uncovered": 2,
  "recovery_plans": [
    { "failure": "gps_loss", "policy": "GpsRecovery", "actions": ["enter_degraded_mode", "imu_navigate"] }
  ],
  "missing_paths": [
    { "failure": "swarm_member_loss", "recommendation": "Add continuity_policy with succession" }
  ]
}

Integration

Engine	Role
Recovery planner	Maps recovery_policy to failure modes
Assurance resilience	resilience_policy completeness
Continuity	Reassign/takeover as recovery paths
Readiness	Recovery coverage factor in go/no-go
Chaos (platform-maturity)	Validates coverage under injected faults (NEXT)
Safety Coverage	Complementary — safety scenarios vs recovery paths

Implementation

Extends spanda-assurance recovery module — composes RecoveryPlanner and failure classification. No duplicate recovery logic.

See differentiation-roadmap.md · self-healing.md · recovery-policies.md.

This site is open source. Improve this page.