feat(recall): ranking release — recency config, tag-score cap, relevance gate, date-aware ranking (#182, #193, #186, #187, #183, #184, #188)

.github/workflows/ci.yml

@@ -12,9 +12,9 @@ name: CI
 
 on:
     push:
-        branches: [main]
+        branches: [main, develop]
     pull_request:
-        branches: [main]
+        branches: [main, develop]
 
 jobs:
     # Code quality checks

CLAUDE.md

@@ -268,9 +268,11 @@ SEARCH_WEIGHT_RELATION=0.25        # Graph relationship boost
 SEARCH_WEIGHT_TAG=0.20             # Tag overlap scoring
 SEARCH_WEIGHT_EXACT=0.20           # Exact phrase in metadata
 SEARCH_WEIGHT_IMPORTANCE=0.10      # Memory importance score
-SEARCH_WEIGHT_RECENCY=0.10         # Linear decay over 180 days
+SEARCH_WEIGHT_RECENCY=0.10         # Decay per SEARCH_RECENCY_* (default: linear over 180 days)
 SEARCH_WEIGHT_CONFIDENCE=0.05      # Memory confidence score
 SEARCH_WEIGHT_RELEVANCE=0.0        # Consolidation decay relevance (disabled by default)
+SEARCH_WEIGHT_TEMPORAL=0.1         # Relative-recency bonus, only when recency_bias re-rank runs
+RECALL_RECENCY_BIAS=off            # Recency re-rank default: off|on|auto (per-request recency_bias param overrides)
 
 # Memory content limits
 MEMORY_CONTENT_SOFT_LIMIT=500      # Auto-summarization trigger (chars)
@@ -382,7 +384,7 @@ Shows: total count, date range, archived count, memories by month (bar chart), t
 python scripts/browse_memories.py diagnose 2751e70e
 ```
 
-Checks: FalkorDB existence, archived status, Qdrant presence + embedding quality, recency score (180-day decay), simulated relevance score breakdown (decay factor, access factor, relationship factor, importance floor), and current search weight config. Reports issues at `[CRITICAL]`, `[WARNING]`, and `[INFO]` severity levels.
+Checks: FalkorDB existence, archived status, Qdrant presence + embedding quality, recency score (configurable decay, default 180-day), simulated relevance score breakdown (decay factor, access factor, relationship factor, importance floor), and current search weight config. Reports issues at `[CRITICAL]`, `[WARNING]`, and `[INFO]` severity levels.
 ### Recall Quality Lab (`scripts/lab/`)
 - **clone_production.sh** - Clone production FalkorDB + Qdrant data to local Docker for safe testing
 - **create_test_queries.py** - Generate test queries with expected results from local data

README.md

@@ -167,6 +167,8 @@ PORT=8001 python app.py
 
 Requires Python 3.10+ (3.12 recommended). → [INSTALLATION.md](INSTALLATION.md#bare-api-development)
 
+**Contributing:** feature PRs target the `develop` branch. `main` only moves via validated release merges, so users deploying from `main` (e.g. Railway auto-deploys) see one deploy per release instead of one per PR.
+
 ## Connect your AI
 
 | Client | Mode | Setup |

app.py

@@ -317,6 +317,7 @@ def require_api_token() -> None:
     get_openai_client=get_openai_client,
     classification_model=CLASSIFICATION_MODEL,
     logger=logger,
+    stats=state.classification_stats,
 )
 
 

automem/api/enrichment.py

@@ -29,6 +29,7 @@ def enrichment_status() -> Any:
             "inflight": inflight,
             "max_attempts": max_attempts,
             "stats": state.enrichment_stats.to_dict(),
+            "classification": state.classification_stats.to_dict(),
         }
         return jsonify(response)