Research-Stack/6-Documentation/docs/semantics/schema_coverage_report.md

ENE SessionArchive Schema Coverage Report

Generated: 2026-04-18
Source: ExtensionScaffold/ENE/SessionArchive.lean vs imported records
Status: Schema supports types, but migration uses defaults

---

Schema Types (Defined in SessionArchive.lean)

SessionSource (5 values)

Value	Used?	Notes
userEditSession	❌ No	Interactive editing sessions
antigravityCodingSession	❌ No	AG coding workflows
ptosBootSession	✅ Yes	Graph address records
eneSwarmSession	❌ No	ENE swarm runs
archivedResearchSession	✅ Yes	Most imported records

Coverage: 2/5 (40%)

SessionRecordKind (5 values)

Value	Used?	Notes
provenance	❌ No	Origin/lineage records
bootAttestation	✅ Yes	Graph OS manifest
swarmRun	❌ No	ENE swarm executions
researchSession	✅ Yes	JSON catalogs
ingestAttestation	✅ Yes	SQLite packages

Coverage: 3/5 (60%)

ArtifactRole (6 values)

Value	Used?	Notes
created	❌ No	New artifacts
modified	❌ No	Changed artifacts
related	✅ Yes	All records (default)
dependency	❌ No	Required components
empiricalAnchor	❌ No	Hardware/physical refs
codeback	❌ No	Derived code

Coverage: 1/6 (17%) — UNDERUTILIZED

ArtifactType (14 values)

Value	Used?	Notes
rustModule	❌ No	*.rs files
rustBinary	❌ No	Compiled Rust
pythonTest	❌ No	*.py test scripts
jsonSchema	❌ No	Schema definitions
verilog	❌ No	*.v files
vhdl	❌ No	*.vhd files
boardLayout	❌ No	PCB files
shader	❌ No	*.wgsl files
document	❌ No	*.md specs
script	❌ No	Shell scripts
dataFile	✅ Yes	All records (default)
chatSession	❌ No	Chat transcripts
attestation	❌ No	Verification proofs
other	❌ No	Uncategorized

Coverage: 1/14 (7%) — SEVERELY UNDERUTILIZED

---

Gap Analysis

What's Missing in Migration

Source JSON files contain:

• python_script → Should map to ArtifactType.pythonTest
• rust_module → Should map to ArtifactType.rustModule
• specification → Should map to ArtifactType.document
• theorem_document → Should map to ArtifactType.document
• academic_paper → Should map to ArtifactType.document
• video → Should map to ArtifactType.other

SessionArchive has examples of:

• regimeTrackerAndHardeningRecord uses .created, .modified roles
• sovereignStackBootRecord uses .empiricalAnchor role
• eneSwarmRunRecord uses .codeback role
• Various artifact types: .rustModule, .rustBinary, .pythonTest, .jsonSchema

Current Migration vs Schema Capability

Current:   { role := .related, artifactType := .dataFile, ... }
Capacity:  { role := .created/.modified/.dependency/.empiricalAnchor/.codeback,
             artifactType := .rustModule/.pythonTest/.document/.chatSession/... }

---

Recommendations

1. Enhance Migration Script

Update ene_import.py to detect file types and set appropriate ArtifactType:

type_map = {
    '.py': 'pythonTest',
    '.rs': 'rustModule',
    '.md': 'document',
    '.json': 'jsonSchema',
    '.toml': 'other',
}

2. Add Role Inference

Set ArtifactRole based on context:

• If file was imported from catalog → .related
• If file was created in session → .created
• If file is a dependency → .dependency

3. Add Missing SessionSource/Kind Combinations

If importing:

• Chat session transcripts → SessionSource.eneSwarmSession
• Code generation outputs → SessionSource.antigravityCodingSession

---

Schema Extension Needed?

No. The existing schema in SessionArchive.lean covers all observed types:

• 5 SessionSource values
• 5 SessionRecordKind values
• 6 ArtifactRole values
• 14 ArtifactType values

The issue is migration fidelity, not schema coverage.

---

Files

File	Purpose
SessionArchive.lean	Schema definition (complete)
ene_import.py	Migration script (needs enhancement)
AutoImported.lean	Generated records (uses defaults)

---

Conclusion

✅ Schema is sufficient — All legacy database types can be represented
⚠️ Migration needs work — Currently uses dataFile/related for everything
💡 Enhancement path clear — Add type detection to ene_import.py