From bad41d63132a5bd00a50056d44ad974736ce6cc1 Mon Sep 17 00:00:00 2001
From: Matteo Cherubini <cherubinimatte@gmail.com>
Date: Wed, 10 Jun 2026 17:20:02 +0200
Subject: [PATCH] feat: Introduce `cross_source` flag for genome registry
 entries

---
 registry.sh              | 12 +++++++-----
 scripts/add-genome.sh    | 18 ++++++++++++++----
 scripts/setup-genomes.sh |  7 ++++---
 3 files changed, 25 insertions(+), 12 deletions(-)

diff --git a/registry.sh b/registry.sh
index 596c462..2c20200 100644
--- a/registry.sh
+++ b/registry.sh
@@ -19,13 +19,15 @@ LIB_DIR="${PROJECT_ROOT}/lib"
 PROVIDERS_DIR="${PROJECT_ROOT}/providers"
 
 # --- GENOME REGISTRY ---
-# Format: "name|description|linked_repo"
-#   - linked_repo is OPTIONAL. Leave empty (trailing pipe) for knowledge-only genomes.
+# Format: "name|description|linked_repo|cross_source"
+#   - linked_repo: OPTIONAL. Leave empty for knowledge-only genomes.
+#   - cross_source: "yes" or "no" (default: no). Controls whether the collector
+#     may read this genome as a source during cross-genome pulls.
 #
 # HOW TO CUSTOMIZE:
 # Replace the placeholder below with your actual genome domains.
-# Example: "genome-work|Work notes and architecture logs|"
-#          "genome-finance|Personal finance|user/repo-finance"
+# Example: "genome-work|Work notes and architecture logs||no"
+#          "genome-finance|Personal finance|user/repo-finance|no"
 GENOMES=(
-  "genome-example|Template genome description for knowledge management|"
+  "genome-example|Template genome description for knowledge management||no"
 )
diff --git a/scripts/add-genome.sh b/scripts/add-genome.sh
index dc4dd6c..0bf6b54 100644
--- a/scripts/add-genome.sh
+++ b/scripts/add-genome.sh
@@ -11,18 +11,28 @@ source "registry.sh"
 
 GENOME_NAME="${1:-}"
 GENOME_DESC="${2:-}"
-GENOME_LINKED="${3:-}"   # optional: linked project repo reference
+GENOME_LINKED="${3:-}"      # optional: linked project repo reference
+GENOME_CROSS_SOURCE="${4:-no}" # optional: cross_source flag (default: no)
 
+# 1. Check mandatory arguments first
 if [[ -z "$GENOME_NAME" || -z "$GENOME_DESC" ]]; then
     error "Missing arguments."
-    echo "Usage: $0 <genome-name> <description> [linked-repo]"
+    echo "Usage: $0 <genome-name> <description> [linked-repo] [cross_source]"
+    echo "  cross_source: yes|no (default: no)"
+    exit 1
+fi
+
+# 2. Then validate the flag if a non-default value was passed
+if [[ "$GENOME_CROSS_SOURCE" != "yes" && "$GENOME_CROSS_SOURCE" != "no" ]]; then
+    error "Invalid cross_source value: $GENOME_CROSS_SOURCE"
+    echo "cross_source must be 'yes' or 'no'"
     exit 1
 fi
 
 step "Adding New Genome: ${GENOME_NAME}"
 
-# Build a 3-field registry entry (linked_repo may be empty)
-GENOMES=("${GENOME_NAME}|${GENOME_DESC}|${GENOME_LINKED}")
+# Build a 4-field registry entry (linked_repo may be empty, cross_source defaults to no)
+GENOMES=("${GENOME_NAME}|${GENOME_DESC}|${GENOME_LINKED}|${GENOME_CROSS_SOURCE}")
 
 # NOTE — Maintenance smell
 # We source setup-genomes.sh as a library/orchestrator hybrid. This works because:
diff --git a/scripts/setup-genomes.sh b/scripts/setup-genomes.sh
index b30af18..f3aa6cd 100644
--- a/scripts/setup-genomes.sh
+++ b/scripts/setup-genomes.sh
@@ -20,10 +20,11 @@ step "Processing Genome Registry"
 
 for entry in "${GENOMES[@]}"; do
     # 3-field format: name|description|linked_repo (linked_repo optional → may be empty)
-    IFS='|' read -r GENOME_NAME GENOME_DESC GENOME_LINKED <<< "$entry"
-    export GENOME_NAME GENOME_DESC GENOME_LINKED
+    IFS='|' read -r GENOME_NAME GENOME_DESC GENOME_LINKED GENOME_CROSS_SOURCE <<< "$entry"
+    GENOME_CROSS_SOURCE="${GENOME_CROSS_SOURCE:-no}"
+    export GENOME_NAME GENOME_DESC GENOME_LINKED GENOME_CROSS_SOURCE
 
-    info "Processing: ${GENOME_NAME}..."
+    info "Processing: ${GENOME_NAME} (cross_source: ${GENOME_CROSS_SOURCE})..."
 
     # 1. Remote Creation (Idempotent)
     provider_create_repo "${GENOME_NAME}" "${GENOME_DESC}" "true"