Promote yoga benchmark and refresh oracle packet priority

README.md

@@ -4,9 +4,9 @@
 
 [![License: MIT](https://img.shields.io/badge/license-MIT-green)](LICENSE)
 [![Python 3.11+](https://img.shields.io/badge/python-3.11+-blue)](https://www.python.org/)
-[![Techniques](https://img.shields.io/badge/techniques-79-blueviolet)](references/technique_registry.json)
-[![Covered](https://img.shields.io/badge/covered-70-green)](references/technique_registry.json)
-[![Complete](https://img.shields.io/badge/complete-9-brightgreen)](references/technique_registry.json)
+[![Techniques](https://img.shields.io/badge/techniques-89-blueviolet)](references/technique_registry.json)
+[![Covered](https://img.shields.io/badge/covered-79-green)](references/technique_registry.json)
+[![Complete](https://img.shields.io/badge/complete-10-brightgreen)](references/technique_registry.json)
 [![Partial](https://img.shields.io/badge/partial-0-lightgrey)](references/technique_registry.json)
 
 ---
@@ -31,7 +31,7 @@
 This is a **Vedic (Jyotish) astrology analysis system** designed for deep, auditable full-chart readings. It is NOT a simple ephemeris calculator — it is a multi-stage interpretive pipeline that:
 
 1. **Computes** divisional charts (D1/D9/D10/...) via Swiss Ephemeris
-2. **Runs** 79 registered techniques (Dashas, Yogas, Shadbala, Ashtakavarga, Transits...)
+2. **Runs** 89 registered techniques (Dashas, Yogas, Shadbala, Ashtakavarga, Transits...)
 3. **Routes** the analysis through strict workflow paths depending on question type (career / relationship / wealth / timing)
 4. **Audits** every technique used — declaring what was called, what is complete/covered, and which limitations affect confidence
 5. **Degrades gracefully** — limitations are labeled, not silently over-promising
@@ -45,7 +45,7 @@ This is a **Vedic (Jyotish) astrology analysis system** designed for deep, audit
 | Technique Audit Table (confidence labeling) | ✅ | ❌ | ❌ | ❌ |
 | Capability degradation (limits are explicit) | ✅ | ❌ | ❌ | ❌ |
 | MEVG external verification gates | ✅ | ❌ | ❌ | ❌ |
-| 79 techniques registered | ✅ | ✅ (50+) | ✅ (200+) | ✅ |
+| 89 techniques registered | ✅ | ✅ (50+) | ✅ (200+) | ✅ |
 | Traditional algorithm benchmarked | ✅ mixed depth | ✅ | ✅ | ✅ |
 | Docker / MCP Server | ✅ | ❌ | ✅ | ❌ |
 | English docs / PyPI package | ✅ in progress | ✅ | ✅ | ✅ |
@@ -442,7 +442,11 @@ The AI does NOT require the user to name techniques (e.g., "Chara Dasha"). It au
 
 ## Technique Coverage
 
-Current count: **68 techniques** (58 covered, 10 complete, 0 partial, 0 missing)
+Current registry count: **89 techniques** (79 covered, 10 complete, 0 partial, 0 missing).
+
+The table below lists representative high-value entries. Treat
+`references/technique_registry.json` as the source of truth for the full
+machine-readable registry.
 
 | Technique | Status | Notes |
 |-----------|--------|-------|

docs/research/ACTIVE_FRONTS.md

@@ -7,6 +7,7 @@ This file is the small index for the current engineering fronts that still drive
 - Before touching strict workflow or adjudicator logic, check:
   - `/Users/wuyongnaren/Documents/印度占星/docs/research/high_value_fragment_source_map_2026_06_28.md`
   - `/Users/wuyongnaren/Documents/印度占星/docs/research/worktree_fragment_triage_2026_06_28.md`
+  - `/Users/wuyongnaren/Documents/印度占星/docs/research/yoga_benchmark_case_promotion_audit_2026_06_28.md`
 - Treat repo truth as authoritative.
 - Treat `docs/research/local_drafts`, Gemini brain notes, Codex attachments, and WorkBuddy copies as candidate sources only.
 

docs/research/worktree_fragment_triage_2026_06_28.md

@@ -56,8 +56,8 @@ Only promote this report later if a stable human-facing archive is needed.
   - Disposition: useful boundary values were promoted into `/Users/wuyongnaren/Documents/印度占星/references/oracle/dasha_shadbala_oracle_cases.json` as `template_bv_raman_vimshottari_boundary_series`.
   - Follow-up: collect source artifact metadata before promoting beyond `template_only`.
 - `/Users/wuyongnaren/Documents/印度占星/tests/test_yoga_benchmark_cases.py`
-  - Current state: smoke-style test ending in `assert True`.
-  - Better path: replace with concrete yoga names and expected detections before promotion.
+  - Disposition: promoted into a real benchmark with concrete `rule_id` expectations and metadata contract checks.
+  - See `/Users/wuyongnaren/Documents/印度占星/docs/research/yoga_benchmark_case_promotion_audit_2026_06_28.md`.
 
 ### Superseded or Needs Review
 

docs/research/yoga_benchmark_case_promotion_audit_2026_06_28.md

@@ -0,0 +1,40 @@
+# Yoga Benchmark Case Promotion Audit - 2026-06-28
+
+## Scope
+
+This pass converts the untracked `tests/test_yoga_benchmark_cases.py` scratch smoke test into an executable Yoga benchmark.
+
+## Problem With The Old Fragment
+
+The old file only verified that `detect_yogas_from_json()` returned a list and ended with `assert True`. It did not constrain any specific Yoga rule, category, strength, or metadata.
+
+## Promoted Test Contract
+
+The promoted test uses the same B.V. Raman-style static Aquarius chart but now asserts stable rule ids:
+
+- `dharma_karmadhipati`
+- `chandra_yoga_exalted`
+- `shakti_yoga_9_10_lord`
+- `bvr_016_vesi_precise`
+- `budha_shukra_yoga`
+- `kalatra_yoga_venus_kendra`
+- `raja_yoga_5_10_lord`
+- `raja_yoga_4_10_lord`
+
+It also verifies that the metadata contract remains useful:
+
+- `category`
+- `strength`
+- `source`
+- human-readable `combination`
+
+## Boundary
+
+- This is a deterministic engine benchmark, not a life prediction.
+- It does not claim B.V. Raman biography validation.
+- It guards the data-driven Yoga engine and rule metadata against silent drift.
+
+## Regression Coverage
+
+- `/Users/wuyongnaren/Documents/印度占星/tests/test_yoga_benchmark_cases.py`
+- `/Users/wuyongnaren/Documents/印度占星/tests/test_yoga_rules_integrity.py`

references/skill_gap_truth_registry.json

@@ -132,7 +132,7 @@
       "id": "covered_is_not_complete",
       "wrong_pattern": "Using covered and complete as if they meant the same thing.",
       "corrected_truth": "Covered means there is a stable entry or usable layer; complete requires mature workflow, tests, boundary text and, where needed, external oracle closure.",
-      "source_ref": "docs/research/current_skill_core_gap_rerank_2026_06_26.md"
+      "source_ref": "docs/research/local_drafts/2026-06/current_skill_core_gap_rerank_2026_06_26.md"
     },
     {
       "id": "single_factor_reading_risk",
@@ -150,7 +150,7 @@
       "id": "jaimini_kp_tajika_not_absent",
       "wrong_pattern": "Describing Chara Dasha, KP/Prashna or Tajika/Sahams as absent.",
       "corrected_truth": "These fronts have usable structures; the hard work is tail-difference closure, traditional workflow deepening and external sample validation.",
-      "source_ref": "docs/research/three_fronts_skill_depth_audit_2026_06_26.md"
+      "source_ref": "docs/research/local_drafts/2026-06/three_fronts_skill_depth_audit_2026_06_26.md"
     }
   ],
   "must_not_overclaim": [

scripts/prepare_oracle_capture_packets.py

@@ -19,9 +19,6 @@ from oracle_collection_queue import build_queue, write_evidence_packets  # noqa:
 from oracle_evidence_validator import build_report  # noqa: E402
 
 
-FIRST_PRIORITY_CAPTURE_ID = "external_template_steve_jobs_dasha_lahiri"
-
-
 def _resolve(path: str) -> Path:
     candidate = Path(path)
     if candidate.is_absolute():
@@ -41,6 +38,26 @@ def _write_json(path: Path, value: dict[str, Any]) -> None:
         fh.write("\n")
 
 
+def _first_priority_packet(queue: dict[str, Any], validator_report: dict[str, Any], output_path: Path) -> Path:
+    by_capture_id = {
+        packet.get("capture_id"): packet
+        for task in queue.get("tasks", [])
+        if isinstance((packet := task.get("evidence_packet", {})), dict)
+    }
+    for packet_result in validator_report.get("packets", []):
+        if packet_result.get("valid") is True:
+            continue
+        capture_id = packet_result.get("capture_id")
+        if not capture_id or capture_id not in by_capture_id:
+            continue
+        return output_path / f"{capture_id}.json"
+
+    packets = sorted(path for path in output_path.glob("external_*.json"))
+    if not packets:
+        return output_path / "external_oracle_packet.json"
+    return packets[0]
+
+
 def _write_next_steps(path: Path, oracle_file: str, output_dir: str, first_packet: str) -> None:
     text = f"""# External Oracle Capture Next Steps
 
@@ -53,8 +70,8 @@ First priority packet:
 1. Open JHora, PyJHora, VedAstro, or another documented external source.
 2. Use the exact birth data, ayanamsa, node mode, timezone, and settings in the packet.
 3. Save a redacted screenshot or stdout snippet under `references/oracle/artifacts/`.
-4. Fill all metadata fields and all `target_placeholders`.
-5. Set `status` to `external_verified` only after the artifact and target values are filled.
+4. Fill missing metadata fields and missing `target_placeholders`.
+5. Set or keep `status` as `external_verified` only after the artifact and target values are filled.
 
 不得把本仓库本地输出当作 external oracle。
 
@@ -78,7 +95,7 @@ python3 scripts/oracle_evidence_validator.py \\
   --queue-file /tmp/jyotish_oracle_queue_filled.json
 ```
 
-Draft packets in `{output_dir}` must remain `valid_packets: 0` until real external evidence is filled.
+Packets in `{output_dir}` must not become valid until real external evidence is filled.
 """
     path.write_text(text, encoding="utf-8")
 
@@ -92,8 +109,7 @@ def prepare_packets(oracle_file: str, output_dir: str) -> dict[str, Any]:
     validator_report = build_report(queue)
 
     packets = sorted(path.name for path in output_path.glob("external_*.json"))
-    first_priority = output_path / f"{FIRST_PRIORITY_CAPTURE_ID}.json"
-    first_priority_path = first_priority if first_priority.exists() else output_path / packets[0]
+    first_priority_path = _first_priority_packet(queue, validator_report, output_path)
 
     manifest = {
         "scope": "external_oracle_capture_packet_manifest",

tests/test_oracle_closure_master_dashboard.py

@@ -48,12 +48,12 @@ def test_oracle_closure_master_dashboard_aggregates_all_hard_fronts() -> None:
     assert report["fronts"]["dasha"]["external_verified_tasks"] == 3
     assert report["fronts"]["dasha"]["first_priority"] is None
     assert report["fronts"]["shadbala"]["external_verified_tasks"] == 4
-    assert report["fronts"]["shadbala"]["first_priority"]["case_id"] == "template_redacted_place_shadbala_raman"
+    assert report["fronts"]["shadbala"]["open_tasks"] == 0
+    assert report["fronts"]["shadbala"]["first_priority"] is None
     assert report["fronts"]["tajika_sahams"]["external_verified_tasks"] == 1
     assert report["fronts"]["tajika_sahams"]["open_tasks"] == 4
     assert report["fronts"]["tajika_sahams"]["first_priority"]["case_id"] == "template_einstein_varshaphala_1905_lahiri"
     assert report["fronts"]["tajika_sahams"]["first_priority"]["missing_groups"]["target"]["count"] == 10
-    assert report["fronts"]["shadbala"]["first_priority"]["manual_fill_plan"]["manual_entry_count"] == 55
     assert report["next_action_order"][0]["front"] == "tajika_sahams"
     assert len(report["next_action_order"]) == 1
 
@@ -69,7 +69,7 @@ def test_oracle_closure_master_dashboard_markdown_can_be_written(tmp_path: Path)
     assert "total_tasks: `12`" in markdown
     assert "can_claim_global_oracle_closure: `false`" in markdown
     assert "`dasha` | 3 | 3 | `complete`" in markdown
-    assert "template_redacted_place_shadbala_raman" in markdown
+    assert "`shadbala` | 4 | 4 | `complete`" in markdown
     assert "template_einstein_varshaphala_1905_lahiri" in markdown
     assert "manual entries" in markdown
     assert "metadata missing" in markdown

tests/test_prepare_oracle_capture_packets.py

@@ -32,12 +32,12 @@ def test_prepare_oracle_capture_packets_writes_packets_manifest_and_next_steps(t
     assert completed.returncode == 0, completed.stderr or completed.stdout
     report = json.loads(completed.stdout)
     assert report["scope"] == "external_oracle_capture_packet_preparation"
-    assert report["packet_count"] == 5
-    assert report["validator_summary"]["valid_packets"] == 0
-    assert report["validator_summary"]["ready_for_calibration"] == 0
-    assert report["first_priority_packet"].endswith("external_template_steve_jobs_dasha_lahiri.json")
+    assert report["packet_count"] == 6
+    assert report["validator_summary"]["valid_packets"] == 4
+    assert report["validator_summary"]["ready_for_calibration"] == 4
+    assert report["first_priority_packet"].endswith("external_template_historical_epoch_lahiri.json")
 
-    packet = tmp_path / "external_template_steve_jobs_dasha_lahiri.json"
+    packet = tmp_path / "external_template_historical_epoch_lahiri.json"
     manifest = tmp_path / "capture_manifest.json"
     next_steps = tmp_path / "OPERATOR_NEXT_STEPS.md"
     assert packet.exists()
@@ -45,17 +45,17 @@ def test_prepare_oracle_capture_packets_writes_packets_manifest_and_next_steps(t
     assert next_steps.exists()
 
     packet_data = json.loads(packet.read_text(encoding="utf-8"))
-    assert packet_data["status"] == "draft"
-    assert packet_data["metadata"]["source_artifact"] == "references/oracle/artifacts/"
-    assert packet_data["target_placeholders"]["target.vimshottari_start_date"] is None
+    assert packet_data["status"] == "external_verified"
+    assert packet_data["metadata"]["source_artifact"].startswith("references/oracle/artifacts/")
+    assert packet_data["target_placeholders"]["target.sun_sidereal_longitude_deg"] is None
 
     manifest_data = json.loads(manifest.read_text(encoding="utf-8"))
     assert manifest_data["scope"] == "external_oracle_capture_packet_manifest"
-    assert manifest_data["packet_count"] == 5
+    assert manifest_data["packet_count"] == 6
     assert packet.name in manifest_data["packets"]
 
     guide = next_steps.read_text(encoding="utf-8")
-    assert "external_template_steve_jobs_dasha_lahiri.json" in guide
+    assert "external_template_historical_epoch_lahiri.json" in guide
     assert "oracle_evidence_validator.py" in guide
     assert "--apply-packet" in guide
     assert "不得把本仓库本地输出当作 external oracle" in guide

tests/test_public_benchmark_dashboard.py

@@ -38,7 +38,7 @@ def test_public_benchmark_dashboard_outputs_stable_json_summary() -> None:
     assert report["schema_version"] == 1
     assert report["summary"]["technique_count"] >= 60
     assert report["summary"]["capability_valid"] is True
-    assert report["oracle_readiness"]["total_packets"] == 5
+    assert report["oracle_readiness"]["total_packets"] == 6
     assert report["oracle_readiness"]["valid_packets"] == 4
     assert report["oracle_readiness"]["ready_for_calibration"] == 4
     assert report["oracle_readiness"]["production_tuning_allowed"] is False

tests/test_yoga_benchmark_cases.py

@@ -0,0 +1,60 @@
+#!/usr/bin/env python3
+"""Executable Yoga benchmark cases recovered from older scratch tests."""
+
+from __future__ import annotations
+
+from scripts.yoga_engine import detect_yogas_from_json
+
+
+RAMAN_STATIC_CHART = {
+    "ascendant": "Aquarius",
+    "planets": {
+        "Sun": {"sign": "Cancer", "house": 6, "degree": 23.0},
+        "Moon": {"sign": "Taurus", "house": 4, "degree": 23.5},
+        "Mars": {"sign": "Leo", "house": 7, "degree": 21.0},
+        "Mercury": {"sign": "Leo", "house": 7, "degree": 14.0},
+        "Jupiter": {"sign": "Scorpio", "house": 10, "degree": 13.0},
+        "Venus": {"sign": "Leo", "house": 7, "degree": 2.0},
+        "Saturn": {"sign": "Taurus", "house": 4, "degree": 10.0},
+        "Rahu": {"sign": "Pisces", "house": 2, "degree": 25.0},
+        "Ketu": {"sign": "Virgo", "house": 8, "degree": 25.0},
+    },
+    "context": {
+        "d9": {
+            "ascendant": "Aries",
+        }
+    },
+}
+
+
+def test_raman_static_chart_triggers_stable_yoga_rule_ids() -> None:
+    detected = detect_yogas_from_json(RAMAN_STATIC_CHART)
+    rule_ids = {row.get("rule_id") for row in detected}
+
+    assert len(detected) >= 50
+    assert {
+        "dharma_karmadhipati",
+        "chandra_yoga_exalted",
+        "shakti_yoga_9_10_lord",
+        "bvr_016_vesi_precise",
+        "budha_shukra_yoga",
+        "kalatra_yoga_venus_kendra",
+        "raja_yoga_5_10_lord",
+        "raja_yoga_4_10_lord",
+    }.issubset(rule_ids)
+
+
+def test_raman_static_chart_preserves_yoga_metadata_contract() -> None:
+    detected = detect_yogas_from_json(RAMAN_STATIC_CHART)
+    by_rule_id = {row.get("rule_id"): row for row in detected}
+
+    dharma = by_rule_id["dharma_karmadhipati"]
+    assert dharma["category"] == "raja"
+    assert dharma["strength"] == "强"
+    assert "9宫主金星Venus" in dharma["combination"]
+    assert "10宫主火星Mars" in dharma["combination"]
+
+    chandra = by_rule_id["chandra_yoga_exalted"]
+    assert chandra["category"] == "chandra"
+    assert chandra["strength"] == "强"
+    assert chandra["source"] == "BPHS"