Drift‑bounded test suite for archive_org operators#

(drop‑in for /docs/archive_org/RTTcodes/tests.md)

1. Test suite overview#

Goal: verify that all six archive_org operators:

• respect drift bounds
• preserve continuity
• honor substrate constraints
• follow the canonical operator order

Scope:

• unit‑style tests per operator
• chain‑style tests for full workflow
• negative tests for unsafe behavior

2. Test matrix#

3. Unit tests (per operator)#

T1 — METADATA_OPERATOR: normalize IA metadata#

Input (pseudo‑RTTcode):

{
  "target": "https://archive.org/details/government-documents"
}

Expected:

• substrate_type = "collection"
• regime = "institutional"
• drift_sensitivity in ["low","medium","high"] (non‑empty)
• coherence non‑empty
• lineage_ids is an array (may be length ≥ 1)

Failure if:

• any of these fields missing or empty
• substrate_type not "collection" for a collection URL

T2 — WAYBACK_OPERATOR: snapshots + drift#

Input:

• target = same URL
• constraints.max_snapshots = 5

Expected:

• snapshots.length ≥ 1
• each snapshot has timestamp + uri
• drift_map present (may be empty object)
• continuity_breaks present (array)

Failure if:

• snapshots missing or malformed
• drift_map missing

T3 — LINEAGE_OPERATOR: continuity kernel#

Input:

• snapshots from T2
• metadata from T1

Expected:

• lineage_graph present
• transformations array present
• regime_shifts array present

Failure if:

• lineage_graph empty while snapshots > 1

T4 — COLLECTION_OPERATOR: envelope#

Input:

• target = collection URL

Expected:

• collection_id non‑empty
• coherence_clusters array present
• regime_profile object present

T5 — PRESERVATION_OPERATOR: substrate#

Input:

• target = collection URL

Expected:

• format non‑empty
• stability_score between 0 and 1
• drift_risk in ["low","medium","high"]

T6 — DRIFTBOUND_RETRIEVAL_OPERATOR: synthesis#

Input:

• context containing outputs from T1–T5 + goal

Expected:

• answer.summary non‑empty
• warnings array present
• if any upstream drift is non‑zero → at least one warning or explicit “no drift” statement

Failure if:

• upstream drift exists but warnings is empty

4. Chain tests (full workflow)#

C1 — Stable collection (Government Documents)#

Target:
https://archive.org/details/government-documents

Expected:

• METADATA: drift_sensitivity = "low"
• WAYBACK: drift_map mostly "none" or "minor"
• LINEAGE: no major regime shifts
• PRESERVATION: stability_score ≥ 0.9
• DRIFTBOUND: summary states “stable” or equivalent, warnings empty

C2 — Moderate drift (Vintage Software)#

Target:
https://archive.org/details/vintagesoftware

Expected:

• WAYBACK: at least one "moderate" drift segment
• LINEAGE: at least one regime shift
• PRESERVATION: stability_score between 0.6 and 0.85
• DRIFTBOUND:
  • summary mentions drift / format changes
  • warnings includes at least one drift warning

C3 — High drift emphasis (synthetic)#

Use a synthetic RTTcode context where:

• drift_map has "high" segments
• continuity_breaks non‑empty
• stability_score ≤ 0.5

Expected:

• DRIFTBOUND:
  • summary explicitly warns about unreliability
  • warnings.length ≥ 1
  • no claim of “stable” or “high confidence”

5. Negative tests (safety boundaries)#

N1 — No direct content reasoning#

Scenario:

• Call DRIFTBOUND_RETRIEVAL_OPERATOR with context missing METADATA / WAYBACK / LINEAGE / PRESERVATION outputs.

Expected:

• operator must refuse or mark result as invalid
• test passes only if:
  • an error is raised, or
  • warnings includes “insufficient operator context”

N2 — Missing substrate profile#

Scenario:

• Provide full context except PRESERVATION_OPERATOR output.

Expected:

• DRIFTBOUND must not claim high reliability
• must include a warning about missing substrate information

6. How to run these tests (conceptually)#

For each operator:

1. Construct a minimal RTTcode input.
2. Run the operator implementation.
3. Validate the output against the expectations above.

For the full chain:

1. Start from target + goal.

2. Run operators in canonical order.

3. Check that:

   • all required fields are present
   • drift is surfaced
   • continuity is explicit
   • substrate is respected
   • final answer matches expectations for the target

This suite doesn’t prescribe a specific test runner—
it defines the behavioral contract that any implementation of the archive_org operators must satisfy.