Implement Roadmap-First Retrieval and Enhance Planning AI Features

- Introduced a roadmap-first approach for retrieval, allowing for structured exercise suggestions based on stage specifications and major steps.
- Added functionality to generate gap-fill offers for unfilled roadmap stages, improving the relevance of exercise recommendations.
- Updated the `ExerciseProgressionPathBuilder` to support the new roadmap-first feature, enhancing user experience with clearer exercise paths.
- Incremented application version to 0.8.206 and updated the database schema version to reflect these changes.

.claude/docs/working/PLANNING_EXERCISE_SUGGEST_CONTEXT.md

@@ -506,7 +506,7 @@ Nach Pfad-Bildung:
 | Prompts | Migration **078/079** — Slugs in `ai_prompts` (Admin), **kein** Template im Python-Code |
 | UI | `ExerciseProgressionPathBuilder` — Roadmap-Box (Major Steps) |
 
-**Übergang:** `include_roadmap_preview=true` liefert `progression_roadmap` **parallel** zum retrieval-first Pfad. **Ziel F3:** `roadmap_first=true` steuert Retrieval.
+**F3 (0.8.206):** `roadmap_first=true` (Default im UI) — Retrieval pro `stage_spec`/Major Step; `roadmap_unfilled` Gap-Angebote. Ohne Flag: retrieval-first wie bisher, Roadmap nur Preview.
 
 **Mitai Workflow-Engine:** bewusst **nicht** jetzt — Pipeline workflow-ready für spätere Anbindung.
 

.claude/docs/working/PLANNING_PROGRESSION_ROADMAP_SPEC.md

@@ -187,7 +187,7 @@ Jede Phase: `(ctx) → ctx`, Zwischenergebnisse in API-Response für **Human-in-
 | **F0** | Spec + Doku + `planning_progression_roadmap.py` Scaffold | 🔄 0.8.204 |
 | **F1** | `include_roadmap_preview` in API + deterministische A/B | 🔄 0.8.204 |
 | **F2** | LLM Phase A/B/C über `ai_prompts` (078/079), `include_llm_roadmap` | 🔄 0.8.205 |
-| **F3** | Retrieval aus `stage_specs` (roadmap_first) | 🔲 |
+| **F3** | Retrieval aus `stage_specs` (roadmap_first) | ✅ 0.8.206 |
 | **F4** | UI Roadmap-Review | 🔲 |
 | **F5** | Trainingsplanung: eigene Pipeline + ggf. Workflow-Engine | 🔲 |
 

backend/planning_exercise_path_builder.py

@@ -23,7 +23,11 @@ from planning_exercise_path_qa import (
     strip_off_topic_steps_from_path,
     try_llm_qa_progression_path,
 )
-from planning_exercise_path_ai_fill import apply_gap_fill_after_qa, collect_gap_fill_specs
+from planning_exercise_path_ai_fill import (
+    apply_gap_fill_after_qa,
+    build_gap_fill_offer,
+    collect_gap_fill_specs,
+)
 from planning_exercise_retrieval import run_multistage_planning_retrieval
 from planning_exercise_semantics import (
     PlanningSemanticBrief,
@@ -47,8 +51,14 @@ from planning_exercise_suggest import (
     resolve_planning_exercise_intent,
 )
 from planning_progression_roadmap import (
+    MajorStep,
+    ProgressionRoadmapContext,
+    StageSpecArtifact,
+    build_roadmap_unfilled_gap_specs,
     progression_roadmap_to_api_dict,
+    resolve_step_exercise_kind_filter,
     run_progression_roadmap_pipeline,
+    stage_spec_retrieval_query,
 )
 from routers.training_planning import _has_planning_role
 
@@ -169,8 +179,12 @@ def _run_path_step_retrieval(
     step_b: Optional[Dict[str, Any]] = None,
     path_target_profile: Optional[PlanningTargetProfile] = None,
     path_intent: Optional[str] = None,
+    step_query_override: Optional[str] = None,
+    step_phase_override: Optional[str] = None,
 ) -> Tuple[List[Dict[str, Any]], PlanningTargetProfile, Dict[str, Any], str]:
-    step_query = step_retrieval_query(semantic_brief, goal_query, step_index, max_steps)
+    step_query = step_query_override or step_retrieval_query(
+        semantic_brief, goal_query, step_index, max_steps
+    )
     if bridge_mode and step_a and step_b:
         phase = step_phase_for_index(semantic_brief, step_index, max_steps)
         parts = [semantic_brief.primary_topic or semantic_brief.retrieval_query or goal_query]
@@ -200,7 +214,8 @@ def _run_path_step_retrieval(
         "has_planning_reference": bool(planned_ids or anchor_id or bridge_mode),
         "semantic_brief": semantic_brief,
         "retrieval_query": step_query,
-        "path_step_phase": step_phase_for_index(semantic_brief, step_index, max_steps),
+        "path_step_phase": step_phase_override
+        or step_phase_for_index(semantic_brief, step_index, max_steps),
     }
     pack = apply_progression_context_to_pack(
         cur,
@@ -331,6 +346,130 @@ def _make_bridge_search_fn(
     return _bridge_search
 
 
+def _annotate_roadmap_step(
+    step: Dict[str, Any],
+    *,
+    stage_spec: StageSpecArtifact,
+    major_step: Optional[MajorStep],
+) -> Dict[str, Any]:
+    reasons = list(step.get("reasons") or [])
+    learning_goal = (stage_spec.learning_goal or "").strip()
+    if learning_goal:
+        roadmap_reason = f"Roadmap: {learning_goal[:120]}"
+        if roadmap_reason not in reasons:
+            reasons.insert(0, roadmap_reason)
+    step["reasons"] = reasons[:4]
+    step["roadmap_major_step_index"] = stage_spec.major_step_index
+    step["roadmap_phase"] = major_step.phase if major_step else None
+    step["roadmap_learning_goal"] = learning_goal or None
+    step["roadmap_match_source"] = "stage_spec"
+    return step
+
+
+def _build_steps_roadmap_first(
+    cur,
+    *,
+    tenant: TenantContext,
+    body: ProgressionPathSuggestRequest,
+    goal_query: str,
+    max_steps: int,
+    semantic_brief: PlanningSemanticBrief,
+    path_target_profile: PlanningTargetProfile,
+    path_intent: str,
+    roadmap_ctx: ProgressionRoadmapContext,
+) -> Tuple[List[Dict[str, Any]], List[Tuple[int, StageSpecArtifact]]]:
+    """Retrieval pro stage_spec statt iterativem Pfad-Bau (Phase F3)."""
+    stage_specs = list(roadmap_ctx.stage_specs or [])[:max_steps]
+    if not stage_specs and roadmap_ctx.roadmap:
+        stage_specs = [
+            StageSpecArtifact(
+                major_step_index=m.index,
+                learning_goal=m.learning_goal,
+            )
+            for m in roadmap_ctx.roadmap.major_steps[:max_steps]
+        ]
+
+    major_by_index: Dict[int, MajorStep] = {}
+    if roadmap_ctx.roadmap:
+        major_by_index = {m.index: m for m in roadmap_ctx.roadmap.major_steps}
+
+    used: Set[int] = set()
+    steps: List[Dict[str, Any]] = []
+    planned_ids: List[int] = []
+    anchor_id: Optional[int] = None
+    anchor_variant_id: Optional[int] = None
+    unfilled: List[Tuple[int, StageSpecArtifact]] = []
+
+    for step_index, stage_spec in enumerate(stage_specs):
+        major = major_by_index.get(stage_spec.major_step_index)
+        step_query = stage_spec_retrieval_query(
+            semantic_brief=semantic_brief,
+            goal_query=goal_query,
+            stage_spec=stage_spec,
+            major_step=major,
+        )
+        step_kind = resolve_step_exercise_kind_filter(stage_spec, body.exercise_kind_any)
+
+        hits, _, _, _ = _run_path_step_retrieval(
+            cur,
+            tenant=tenant,
+            goal_query=goal_query,
+            step_index=step_index,
+            max_steps=max_steps,
+            planned_ids=planned_ids,
+            anchor_id=anchor_id,
+            anchor_variant_id=anchor_variant_id,
+            progression_graph_id=body.progression_graph_id,
+            include_llm_intent=body.include_llm_intent and step_index == 0,
+            exercise_kind_any=step_kind,
+            semantic_brief=semantic_brief,
+            path_target_profile=path_target_profile,
+            path_intent=path_intent,
+            step_query_override=step_query,
+            step_phase_override=major.phase if major else None,
+        )
+
+        hit = _pick_best_path_hit(hits, used, semantic_brief=semantic_brief)
+        if not hit and step_query != goal_query:
+            hits, _, _, _ = _run_path_step_retrieval(
+                cur,
+                tenant=tenant,
+                goal_query=goal_query,
+                step_index=step_index,
+                max_steps=max_steps,
+                planned_ids=planned_ids,
+                anchor_id=anchor_id,
+                anchor_variant_id=anchor_variant_id,
+                progression_graph_id=body.progression_graph_id,
+                include_llm_intent=False,
+                exercise_kind_any=step_kind,
+                semantic_brief=semantic_brief,
+                path_target_profile=path_target_profile,
+                path_intent=path_intent,
+                step_query_override=goal_query,
+                step_phase_override=major.phase if major else None,
+            )
+            hit = _pick_best_path_hit(hits, used, semantic_brief=semantic_brief)
+
+        if not hit:
+            unfilled.append((step_index, stage_spec))
+            continue
+
+        step = _annotate_roadmap_step(
+            _hit_to_path_step(hit),
+            stage_spec=stage_spec,
+            major_step=major,
+        )
+        steps.append(step)
+        eid = int(step["exercise_id"])
+        used.add(eid)
+        planned_ids.append(eid)
+        anchor_id = eid
+        anchor_variant_id = step.get("variant_id")
+
+    return steps, unfilled
+
+
 def suggest_progression_path(
     cur,
     *,
@@ -360,12 +499,69 @@ def suggest_progression_path(
         include_llm_intent=body.include_llm_intent,
     )
 
+    roadmap_first = bool(body.roadmap_first)
+    include_roadmap = roadmap_first or body.include_roadmap_preview
+    progression_roadmap: Optional[Dict[str, Any]] = None
+    roadmap_ctx: Optional[ProgressionRoadmapContext] = None
+    roadmap_unfilled: List[Tuple[int, StageSpecArtifact]] = []
+    roadmap_gap_offers: List[Dict[str, Any]] = []
+
+    if include_roadmap:
+        roadmap_ctx = run_progression_roadmap_pipeline(
+            goal_query,
+            max_steps=max_steps,
+            semantic_brief=semantic_brief,
+            cur=cur,
+            include_llm_roadmap=body.include_llm_roadmap,
+        )
+        progression_roadmap = progression_roadmap_to_api_dict(roadmap_ctx)
+
     used: Set[int] = set()
     steps: List[Dict[str, Any]] = []
     planned_ids: List[int] = []
     anchor_id: Optional[int] = None
     anchor_variant_id: Optional[int] = None
 
+    if roadmap_first and roadmap_ctx is not None:
+        steps, roadmap_unfilled = _build_steps_roadmap_first(
+            cur,
+            tenant=tenant,
+            body=body,
+            goal_query=goal_query,
+            max_steps=max_steps,
+            semantic_brief=semantic_brief,
+            path_target_profile=path_target_profile,
+            path_intent=path_intent,
+            roadmap_ctx=roadmap_ctx,
+        )
+        planned_ids = [int(s["exercise_id"]) for s in steps if s.get("exercise_id") is not None]
+        if planned_ids:
+            anchor_id = planned_ids[-1]
+            anchor_variant_id = steps[-1].get("variant_id")
+        if body.include_ai_gap_fill and roadmap_unfilled:
+            major_by_index = (
+                {m.index: m for m in roadmap_ctx.roadmap.major_steps}
+                if roadmap_ctx.roadmap
+                else {}
+            )
+            roadmap_gap_specs = build_roadmap_unfilled_gap_specs(
+                unfilled_specs=roadmap_unfilled,
+                major_steps_by_index=major_by_index,
+                steps=steps,
+                brief=semantic_brief,
+                goal_query=goal_query,
+            )
+            for spec in roadmap_gap_specs:
+                roadmap_gap_offers.append(
+                    build_gap_fill_offer(
+                        spec=spec,
+                        steps=steps,
+                        goal_query=goal_query,
+                        brief=semantic_brief,
+                        proposal=None,
+                    )
+                )
+    else:
         for step_index in range(max_steps):
             hits, _tp, _qis, _intent = _run_path_step_retrieval(
                 cur,
@@ -490,6 +686,12 @@ def suggest_progression_path(
                 auto_insert_proposals=False,
             )
 
+    if roadmap_gap_offers:
+        seen_offer_ids = {o.get("offer_id") for o in gap_fill_offers}
+        for offer in roadmap_gap_offers:
+            if offer.get("offer_id") not in seen_offer_ids:
+                gap_fill_offers.append(offer)
+
     path_qa = build_path_qa_summary(
         gaps=gaps,
         bridge_inserts=bridge_inserts,
@@ -505,6 +707,8 @@ def suggest_progression_path(
 
     target_profile_summary = path_target_profile.to_summary_dict(cur)
     retrieval_parts = ["profile_v1", "full_library", "path_builder", "semantics"]
+    if roadmap_first:
+        retrieval_parts.append("roadmap_first")
     if body.include_path_qa:
         retrieval_parts.append("path_qa")
     if llm_qa_applied:
@@ -515,20 +719,10 @@ def suggest_progression_path(
         retrieval_parts.append("ai_gap_fill")
     if gap_fill_offers:
         retrieval_parts.append("gap_fill_offers")
-
-    progression_roadmap: Optional[Dict[str, Any]] = None
-    if body.include_roadmap_preview or body.roadmap_first:
-        roadmap_ctx = run_progression_roadmap_pipeline(
-            goal_query,
-            max_steps=max_steps,
-            semantic_brief=semantic_brief,
-            cur=cur,
-            include_llm_roadmap=body.include_llm_roadmap,
-        )
-        progression_roadmap = progression_roadmap_to_api_dict(roadmap_ctx)
+    if include_roadmap:
         retrieval_parts.append("roadmap_preview")
-        if body.roadmap_first:
-            retrieval_parts.append("roadmap_first_pending")
+    if roadmap_unfilled:
+        retrieval_parts.append("roadmap_unfilled")
 
     return {
         "goal_query": goal_query,
@@ -543,7 +737,8 @@ def suggest_progression_path(
         "path_qa": path_qa,
         "gap_fill_offers": gap_fill_offers,
         "progression_roadmap": progression_roadmap,
-        "roadmap_first": body.roadmap_first,
+        "roadmap_first": roadmap_first,
+        "roadmap_unfilled_count": len(roadmap_unfilled),
         "retrieval_phase": "+".join(retrieval_parts),
     }
 

backend/planning_progression_roadmap.py

@@ -13,7 +13,7 @@ from __future__ import annotations
 import json
 import logging
 import re
-from typing import Any, Dict, List, Optional, Sequence, Tuple
+from typing import Any, Dict, List, Mapping, Optional, Sequence, Tuple
 
 from pydantic import BaseModel, Field, ValidationError
 
@@ -421,6 +421,107 @@ def consolidate_micro_to_major(
     return majors, notes
 
 
+def _normalize_query(query: Optional[str]) -> str:
+    return re.sub(r"\s+", " ", (query or "").strip())
+
+
+def stage_spec_retrieval_query(
+    *,
+    semantic_brief: PlanningSemanticBrief,
+    goal_query: str,
+    stage_spec: StageSpecArtifact,
+    major_step: Optional[MajorStep] = None,
+) -> str:
+    """Retrieval-Query für einen Roadmap-Major-Step (Phase F3)."""
+    parts: List[str] = []
+    topic = (semantic_brief.primary_topic or semantic_brief.retrieval_query or goal_query).strip()
+    if topic:
+        parts.append(topic)
+    learning_goal = (stage_spec.learning_goal or "").strip()
+    if learning_goal:
+        parts.append(learning_goal)
+    phase = (major_step.phase if major_step else "").strip().lower()
+    if phase:
+        parts.append(phase)
+    if stage_spec.load_profile:
+        parts.extend(str(x).strip() for x in stage_spec.load_profile[:2] if str(x).strip())
+    exercise_type = (stage_spec.exercise_type or "").strip().lower()
+    if exercise_type == "partner_drill":
+        parts.append("partner")
+    elif exercise_type == "kombination":
+        parts.append("kombination")
+    return _normalize_query(" ".join(parts)) or _normalize_query(goal_query)
+
+
+def stage_spec_exercise_kind_filter(stage_spec: StageSpecArtifact) -> Optional[List[str]]:
+    """Mappt didaktischen exercise_type auf DB exercise_kind (simple/combination)."""
+    et = (stage_spec.exercise_type or "").strip().lower()
+    if et == "kombination":
+        return ["combination"]
+    if et in ("kihon_einzel", "partner_drill", "grundtechnik"):
+        return ["simple"]
+    return None
+
+
+def resolve_step_exercise_kind_filter(
+    stage_spec: StageSpecArtifact,
+    request_filter: Optional[Sequence[str]],
+) -> Optional[List[str]]:
+    """Schnittmenge aus Roadmap-Stufe und optionalem Request-Filter."""
+    stage_filter = stage_spec_exercise_kind_filter(stage_spec)
+    if not request_filter:
+        return stage_filter
+    req = [str(x).strip().lower() for x in request_filter if str(x).strip()]
+    if not stage_filter:
+        return req or None
+    merged = [k for k in stage_filter if k in req]
+    return merged or req
+
+
+def build_roadmap_unfilled_gap_specs(
+    *,
+    unfilled_specs: Sequence[Tuple[int, StageSpecArtifact]],
+    major_steps_by_index: Mapping[int, MajorStep],
+    steps: Sequence[Mapping[str, Any]],
+    brief: PlanningSemanticBrief,
+    goal_query: str,
+) -> List[Dict[str, Any]]:
+    """Gap-Fill-Angebote für Roadmap-Stufen ohne Bibliothekstreffer."""
+    topic = (brief.primary_topic or "Technik").strip()
+    specs: List[Dict[str, Any]] = []
+    for roadmap_idx, stage_spec in unfilled_specs:
+        major = major_steps_by_index.get(stage_spec.major_step_index)
+        phase = (major.phase if major else "vertiefung").strip().lower()
+        insert_after = min(max(roadmap_idx - 1, -1), max(len(steps) - 1, -1))
+        title_hint = (stage_spec.learning_goal or f"{topic} — {phase}").strip()[:120]
+        sketch_parts = [
+            f"Planungsziel: {goal_query}",
+            f"Roadmap-Stufe {stage_spec.major_step_index + 1} ({phase}): {stage_spec.learning_goal}",
+        ]
+        if stage_spec.success_criteria:
+            sketch_parts.append(f"Erfolgskriterien: {', '.join(stage_spec.success_criteria[:3])}")
+        specs.append(
+            {
+                "source": "roadmap_unfilled",
+                "insert_after_index": insert_after,
+                "gap": {
+                    "expected_phase": phase,
+                    "roadmap_major_step_index": stage_spec.major_step_index,
+                    "learning_goal": stage_spec.learning_goal,
+                },
+                "phase": phase,
+                "title_hint": title_hint,
+                "sketch": "\n".join(sketch_parts),
+                "rationale": (
+                    f"Keine passende Bibliotheks-Übung für Roadmap-Stufe "
+                    f"{stage_spec.major_step_index + 1} ({phase})."
+                ),
+                "roadmap_major_step_index": stage_spec.major_step_index,
+            }
+        )
+    return specs[:5]
+
+
 def build_stage_specs(
     major_steps: Sequence[MajorStep],
     *,
@@ -553,7 +654,11 @@ __all__ = [
     "RoadmapArtifact",
     "StageSpecArtifact",
     "build_goal_analysis",
+    "build_roadmap_unfilled_gap_specs",
     "build_stage_specs",
+    "resolve_step_exercise_kind_filter",
+    "stage_spec_exercise_kind_filter",
+    "stage_spec_retrieval_query",
     "consolidate_micro_to_major",
     "develop_micro_objectives",
     "progression_roadmap_to_api_dict",

backend/tests/test_planning_exercise_path_builder.py

@@ -1,5 +1,10 @@
-"""Tests Planungs-KI Phase C3/E — Pfad-Vorschläge."""
-from planning_exercise_path_builder import _pick_best_path_hit, _hit_to_path_step
+"""Tests Planungs-KI Phase C3/E/F — Pfad-Vorschläge."""
+from planning_exercise_path_builder import (
+    _annotate_roadmap_step,
+    _hit_to_path_step,
+    _pick_best_path_hit,
+)
+from planning_progression_roadmap import MajorStep, StageSpecArtifact
 
 
 def test_pick_next_path_hit_skips_used():
@@ -23,3 +28,17 @@ def test_hit_to_path_step_maps_variant():
     assert step["exercise_id"] == 10
     assert step["variant_id"] == 7
     assert step["suggested_variant_name"] == "Leicht"
+
+
+def test_annotate_roadmap_step_adds_metadata():
+    spec = StageSpecArtifact(major_step_index=1, learning_goal="Grundstellung Mae Geri")
+    major = MajorStep(index=1, phase="grundlage", learning_goal=spec.learning_goal, consolidates=["m1"])
+    step = _annotate_roadmap_step(
+        {"exercise_id": 5, "title": "Test", "reasons": ["Bibliothek"]},
+        stage_spec=spec,
+        major_step=major,
+    )
+    assert step["roadmap_major_step_index"] == 1
+    assert step["roadmap_phase"] == "grundlage"
+    assert step["roadmap_match_source"] == "stage_spec"
+    assert any("Roadmap:" in r for r in step["reasons"])

backend/tests/test_planning_progression_roadmap.py

@@ -3,11 +3,17 @@ from planning_progression_roadmap import (
     PROMPT_SLUG_GOAL_ANALYSIS,
     PROMPT_SLUG_ROADMAP,
     PROMPT_SLUG_STAGE_SPEC,
+    MajorStep,
+    StageSpecArtifact,
     build_goal_analysis,
+    build_roadmap_unfilled_gap_specs,
     consolidate_micro_to_major,
     develop_micro_objectives,
     progression_roadmap_to_api_dict,
+    resolve_step_exercise_kind_filter,
     run_progression_roadmap_pipeline,
+    stage_spec_exercise_kind_filter,
+    stage_spec_retrieval_query,
 )
 from planning_exercise_semantics import build_semantic_brief
 
@@ -43,6 +49,47 @@ def test_major_steps_have_learning_goals():
         assert step.consolidates
 
 
+def test_stage_spec_retrieval_query_includes_learning_goal():
+    brief = build_semantic_brief("Mae Geri Perfektion")
+    spec = StageSpecArtifact(
+        major_step_index=1,
+        learning_goal="Koordination und Präzision vertiefen",
+        load_profile=["präzision"],
+        exercise_type="kihon_einzel",
+    )
+    major = MajorStep(index=1, phase="vertiefung", learning_goal=spec.learning_goal, consolidates=["m3"])
+    q = stage_spec_retrieval_query(
+        semantic_brief=brief,
+        goal_query="Mae Geri Perfektion",
+        stage_spec=spec,
+        major_step=major,
+    )
+    assert "vertiefung" in q.lower()
+    assert "Koordination" in q or "Präzision" in q
+
+
+def test_stage_spec_exercise_kind_filter_maps_combination():
+    spec = StageSpecArtifact(major_step_index=0, exercise_type="kombination")
+    assert stage_spec_exercise_kind_filter(spec) == ["combination"]
+    assert resolve_step_exercise_kind_filter(spec, ["simple"]) == ["simple"]
+
+
+def test_build_roadmap_unfilled_gap_specs():
+    brief = build_semantic_brief("Mae Geri")
+    spec = StageSpecArtifact(major_step_index=2, learning_goal="Anwendung im Partnerdrill")
+    major = MajorStep(index=2, phase="anwendung", learning_goal=spec.learning_goal, consolidates=["m5"])
+    specs = build_roadmap_unfilled_gap_specs(
+        unfilled_specs=[(2, spec)],
+        major_steps_by_index={2: major},
+        steps=[{"exercise_id": 1, "title": "A"}, {"exercise_id": 2, "title": "B"}],
+        brief=brief,
+        goal_query="Mae Geri",
+    )
+    assert len(specs) == 1
+    assert specs[0]["source"] == "roadmap_unfilled"
+    assert specs[0]["phase"] == "anwendung"
+
+
 def test_api_dict_exposes_prompt_slug_catalog():
     ctx = run_progression_roadmap_pipeline("Mae Geri", max_steps=3, include_llm_roadmap=False)
     api = progression_roadmap_to_api_dict(ctx)

backend/version.py

@@ -1,6 +1,6 @@
 # Shinkan Jinkendo Version Information
 
-APP_VERSION = "0.8.205"
+APP_VERSION = "0.8.206"
 BUILD_DATE = "2026-06-07"
 DB_SCHEMA_VERSION = "20260606086"
 
@@ -38,7 +38,7 @@ MODULE_VERSIONS = {
     "skill_profiles": "1.0.0",  # Phase 3: gewichtetes Fähigkeiten-Profil + skill-discovery/suggestions
     "methods": "0.1.0",
     "exercises": "2.37.1",  # KI-Endpoints: feature_usage nach ai_calls consume
-    "planning_exercise_suggest": "0.17.1",  # F2: Roadmap-LLM via ai_prompts-Slugs, kein Hardcoding
+    "planning_exercise_suggest": "0.18.0",  # F3: roadmap_first Retrieval pro stage_spec
     "training_units": "0.4.0",  # POST .../publish-to-framework: Ablauf aus geplanter Einheit → Rahmen-Slot-Blueprint
     "training_programs": "0.1.0",
     "planning": "0.15.0",  # Vorlagen: Strukturvorschau, Bearbeiten inkl. Split-Sessions + Beschreibung
@@ -53,6 +53,15 @@ MODULE_VERSIONS = {
 }
 
 CHANGELOG = [
+    {
+        "version": "0.8.206",
+        "date": "2026-06-07",
+        "changes": [
+            "Phase F3: roadmap_first — Bibliotheks-Match pro stage_spec/Major Step statt iterativem Pfad.",
+            "Gap-Angebote für unbesetzte Roadmap-Stufen (roadmap_unfilled).",
+            "UI: Pfad-Builder sendet roadmap_first; Übungen an Roadmap gekoppelt.",
+        ],
+    },
     {
         "version": "0.8.205",
         "date": "2026-06-07",

docs/HANDOVER.md

@@ -1,7 +1,7 @@
 # Shinkan Jinkendo – Entwicklungsstand & Handover
 
 **Stand:** 2026-06-07  
-**App-Version / DB-Schema:** App **`0.8.204`** (Planungs-KI Phase F0); DB **`20260606085`** — maßgeblich **`backend/version.py`**.
+**App-Version / DB-Schema:** App **`0.8.206`** (Planungs-KI Phase F3); DB **`20260606086`** — maßgeblich **`backend/version.py`**.
 
 Diese Datei ist die **Einstiegs-Doku für neue Chat-Sessions**: Anforderungen im Detail stehen in `.claude/docs/` (siehe unten); hier der **implementierte Stand**, **Medien-Meilenstein** und **sinnvolle nächste Schritte**.
 
@@ -107,18 +107,19 @@ Das Schema ist gegenüber dem Code zurück: Migration **`022_skills_schema_compl
 | **E** | Semantik-Schicht (Brief, Phrasen-Score) + Pfad-QA (Lücken, Brücken, LLM-QS) | ✅ **0.8.186** |
 | **E2** | Pfad-Neuordnung (LLM) + KI-Neuanlage bei unüberbrückbaren Lücken | ✅ **0.8.187** |
 | **E3** | `gap_fill_offers`, Off-Topic-Strip, voller KI-Call bei Lücken | ✅ **0.8.203** |
-| **F0–F1** | **Roadmap-first** Progressionsgraph (A→B→C), Workflow-lite, API-Preview | 🔄 **0.8.204** |
+| **F0–F2** | Roadmap-Pipeline + LLM-Prompts (078/079) | ✅ **0.8.205** |
+| **F3** | `roadmap_first` — Retrieval pro `stage_spec` | ✅ **0.8.206** |
 | **D** | `planning_context` an `suggestExerciseAi` (Neu-Anlage) | 🔲 |
 
 **Architektur-Entscheidung (2026-06-07):** Progressionsgraph = **Roadmap-first** (Ziel → Major Steps → Übungs-Match). **Keine Gruppenanalyse** im Graphen. Mitai Workflow-Engine **später** — jetzt `planning_progression_roadmap.py`. Spec: **`.claude/docs/working/PLANNING_PROGRESSION_ROADMAP_SPEC.md`** · Roadmap: **`docs/architecture/PLANNING_KI_ROADMAP.md`**
 
-**Backend:** `planning_exercise_suggest.py`, `planning_exercise_retrieval.py`, `planning_exercise_path_builder.py`, **`planning_progression_roadmap.py`** · Router `POST /api/planning/exercise-suggest`, `POST /api/planning/progression-path-suggest` (`include_roadmap_preview`)
+**Backend:** `planning_exercise_suggest.py`, `planning_exercise_retrieval.py`, `planning_exercise_path_builder.py`, **`planning_progression_roadmap.py`** · Router `POST /api/planning/exercise-suggest`, `POST /api/planning/progression-path-suggest` (`roadmap_first`, `include_roadmap_preview`)
 
-**Frontend:** `ExerciseProgressionPathBuilder` — Roadmap-Box (Major Steps) + bestehender Pfad-Review · `ExercisePickerModal` (Planung)
+**Frontend:** `ExerciseProgressionPathBuilder` — Roadmap-Box + Pfad je Major Step (`roadmap_first`) · `ExercisePickerModal` (Planung)
 
 **Superadmin:** Übungs-Anreicherung (Skills) — `exercise_enrichment_admin` (**0.8.178+**), separater Admin-Flow
 
-**Offen (F2+):** LLM Roadmap (Prompts **078**), `roadmap_first` Retrieval, Roadmap-UI editierbar; Trainingsplanung eigene Pipeline (Gruppenkontext); Enrichment
+**Offen (F4+):** Roadmap-UI editierbar; Trainingsplanung eigene Pipeline (Gruppenkontext); Enrichment
 
 #### Übungs-KI Formular / Schnellanlage (Stand **0.8.171**)
 

docs/architecture/PLANNING_KI_ROADMAP.md

@@ -58,10 +58,11 @@ Diese Roadmap ergänzt die **Architektur-Refaktor-Roadmap** (`UMSETZUNGSPLAN_ROA
 - [x] Deterministischer Fallback wenn Prompt/OpenRouter fehlt
 - [ ] Response/UI: genutzte `prompt_slugs` sichtbar machen (Admin-Hinweis)
 
-### F3 — roadmap-first
+### F3 — roadmap-first (0.8.206)
 
-- [ ] Retrieval pro `major_step` + `stage_spec` statt iterativem Pfad-Bau
-- [ ] QA/Lücken an Roadmap koppeln
+- [x] Retrieval pro `major_step` + `stage_spec` statt iterativem Pfad-Bau
+- [x] Gap-Angebote für unbesetzte Roadmap-Stufen (`roadmap_unfilled`)
+- [ ] QA/Lücken vollständig an Roadmap koppeln (Brücken optional reduzieren)
 
 ### F4 — UI
 

frontend/src/components/ExerciseProgressionPathBuilder.jsx

@@ -58,6 +58,7 @@ const OFFER_SOURCE_LABELS = {
   unfilled_gap: 'Lücke',
   off_topic: 'Themenfremd',
   llm_suggested: 'QS-Empfehlung',
+  roadmap_unfilled: 'Roadmap-Stufe',
 }
 
 function resolveDefaultFocusAreaId(targetSummary, focusAreas) {
@@ -349,6 +350,7 @@ export default function ExerciseProgressionPathBuilder({
         include_ai_gap_fill: true,
         include_roadmap_preview: true,
         include_llm_roadmap: true,
+        roadmap_first: true,
         progression_graph_id: Number(graphId),
       })
       const qa = res?.path_qa || null
@@ -531,7 +533,7 @@ export default function ExerciseProgressionPathBuilder({
             {progressionRoadmap.llm_roadmap_applied
               ? ' (KI-Prompts aus Admin-Konfiguration)'
               : ' (heuristischer Fallback — KI-Prompts in ai_prompts)'}
-            . Übungen unten: Bibliothekssuche (Übergangsphase).
+            . Übungen unten: je Major Step aus der Bibliothek (roadmap-first).
           </p>
           <ol style={{ margin: 0, paddingLeft: '18px', fontSize: '13px', lineHeight: 1.5 }}>
             {progressionRoadmap.roadmap.major_steps.map((step) => (