IB-WP-0019-T07: archive integration; close IB-WP-0019

The default archive include set already pulls output/ in wholesale, so
output/budget/ already lands inside the archive package with no code
change. Add a budget_summary block to ArchiveRecord.metadata so
catalog-level tools can see plans_count, runs_count, total_tokens,
total_cost_usd_known, total_cost_usd_estimated, and the
latest_snapshot_id without unpacking the archive. An infospace with no
budget data still archives cleanly with an empty metadata dict.

Closes IB-WP-0019 (Budget and Usage Registry): T01-T07 all done.
Three-layer design landed end-to-end — layer 1 (per-infospace
plans.yaml / usage.yaml / summary.yaml) and layer 3 (state-hub
record_token_event emission with failure isolation) live here; layer 2
(cross-application QualityLedger for adaptive routing) is parked in
llm-connect LLM-WP-0004 and infospace-bench IB-WP-0018 awaits it.

122 tests pass.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

src/infospace_bench/archive.py

@@ -337,12 +337,65 @@ async def _archive_infospace_async(
         producer=PRODUCER,
         subject=subject,
         store_root=str(effective_store_root) if effective_store_root else None,
+        metadata=_archive_metadata(root),
         skipped_top_level=skipped_top_level,
     )
     _append_index(root, record)
     return record
 
 
+def _archive_metadata(root: Path) -> dict[str, Any]:
+    """Compute a small budget_summary for the archive manifest.
+
+    Lets catalog-level tools find an archived infospace's cost shape without
+    unpacking it. Returns an empty dict when no budget data exists.
+    """
+    from .budget import (
+        read_plan_snapshots,
+        read_run_variance,
+        read_usage_runs,
+    )
+
+    try:
+        plans = read_plan_snapshots(root)
+        runs = read_usage_runs(root)
+        summary = read_run_variance(root)
+    except Exception:
+        return {}
+    if not plans and not runs and summary is None:
+        return {}
+    total_tokens = 0
+    total_cost_known = 0.0
+    total_cost_estimated = 0.0
+    for run in runs:
+        rollup = run.get("rollup") or {}
+        total_tokens += int(rollup.get("total_tokens") or 0)
+        try:
+            total_cost_known += float(rollup.get("total_cost_usd_known") or 0.0)
+        except (TypeError, ValueError):
+            pass
+        estimated = rollup.get("total_cost_usd_estimated")
+        if estimated is not None:
+            try:
+                total_cost_estimated += float(estimated)
+            except (TypeError, ValueError):
+                pass
+    budget_summary: dict[str, Any] = {
+        "plans_count": len(plans),
+        "runs_count": len(runs),
+        "total_tokens": total_tokens,
+        "total_cost_usd_known": round(total_cost_known, 6),
+        "total_cost_usd_estimated": round(total_cost_estimated, 6) if total_cost_estimated else None,
+    }
+    if plans:
+        budget_summary["latest_snapshot_id"] = plans[-1].get("snapshot_id")
+    if summary is not None and isinstance(summary.get("snapshot_id"), str):
+        budget_summary["last_run_snapshot_id"] = summary.get("snapshot_id")
+    return {"budget_summary": budget_summary}
+
+
+
+
 def _collect_files(
     root: Path,
     *,

tests/test_budget_registry.py

@@ -722,6 +722,40 @@ def test_budget_cli_list_and_show(tmp_path: Path) -> None:
     assert show_payload["usage"]["runs"]
 
 
+def test_archive_includes_budget_dir_and_records_summary(tmp_path: Path) -> None:
+    from infospace_bench.archive import archive_infospace
+    from infospace_bench.budget import PLANS_FILE, USAGE_FILE
+    from infospace_bench.generator import run_generation
+
+    root = _build_infospace(tmp_path)
+    fixture = tmp_path / "responses.yaml"
+    _write_minimal_fixture(fixture)
+    plan_generation(root)
+    run_generation(root, fixture_responses=fixture)
+    assert (root / PLANS_FILE).is_file()
+    assert (root / USAGE_FILE).is_file()
+
+    record = archive_infospace(root, retention_class="release-evidence")
+
+    summary = record.metadata.get("budget_summary")
+    assert summary is not None
+    assert summary["plans_count"] >= 1
+    assert summary["runs_count"] == 1
+    assert "latest_snapshot_id" in summary
+    # The budget dir is implicitly included via output/, so the archive's
+    # file_count should reflect that.
+    assert record.file_count > 0
+
+
+def test_archive_metadata_empty_when_no_budget_data(tmp_path: Path) -> None:
+    from infospace_bench.archive import archive_infospace
+
+    root = _build_infospace(tmp_path)
+
+    record = archive_infospace(root, retention_class="release-evidence")
+    assert record.metadata.get("budget_summary") is None or record.metadata == {}
+
+
 def test_plan_cli_writes_snapshot(tmp_path: Path) -> None:
     root = _build_infospace(tmp_path)
     env = os.environ.copy()

workplans/IB-WP-0019-budget-and-usage-registry.md

@@ -4,7 +4,7 @@ type: workplan
 title: "Budget and Usage Registry for Infospaces"
 domain: markitect
 repo: infospace-bench
-status: todo
+status: done
 owner: markitect
 topic_slug: markitect
 created: "2026-05-17"
@@ -195,7 +195,7 @@ state_hub_task_id: "7cb34bfc-c562-4dda-a6d4-b44158644e19"
 
 ```task
 id: IB-WP-0019-T07
-status: todo
+status: done
 priority: low
 state_hub_task_id: "b97906e0-2835-4246-9868-840c02d64fae"
 ```