diff --git a/skills/auto-github-contributor/scripts/browser-verify.sh b/skills/auto-github-contributor/scripts/browser-verify.sh
index 0967223..60f08ea 100755
--- a/skills/auto-github-contributor/scripts/browser-verify.sh
+++ b/skills/auto-github-contributor/scripts/browser-verify.sh
@@ -1,21 +1,23 @@
 #!/usr/bin/env bash
+
 # Browser-based visual verification.
 #
 # This script drives a headless browser to capture screenshots / DOM dumps so
-# the agent can eyeball the change before opening the PR. The actual automation
-# body is intentionally a STUB — wire it up to whichever tool you prefer:
-#
-#   - Playwright    (pnpm dlx playwright screenshot ...)
-#   - Puppeteer     (node ./scripts/puppet.js ...)
-#   - chrome-devtools MCP   (mcp__chrome_devtools__take_screenshot)
-#   - Browser Use MCP
+# the agent can eyeball the change before opening the PR.  The Playwright
+# backend is wired up by default; Puppeteer and chrome-devtools MCP are
+# available as drop-in alternatives (see commented blocks below).
 #
 # Contract:
 #   Input  : --url <url> --out <png-path> [--selector <css>] [--wait <ms>]
 #   Output : writes PNG to --out, prints "SCREENSHOT=<abs-path>" on stdout
-#   Stubs  : when the automation body is not yet implemented, the script
-#            still creates the output directory, writes a placeholder note,
-#            and exits 0 with "stub: <reason>" so the agent flow continues.
+#   Stub   : when the automation body fails, the script still creates the output
+#            directory, writes a placeholder note, and exits 0 so the agent flow
+#            continues.
+#
+# Backend selection (change the single `backend` variable below):
+#   playwright  – Playwright CLI (default, works on all platforms)
+#   puppeteer   – node + puppeteer.mjs
+#   chrome-mcp  – chrome-devtools MCP (agent should use MCP tool directly)
 
 set -euo pipefail
 source "$(dirname "$0")/config.sh"
@@ -24,14 +26,16 @@ URL=""
 OUT=""
 SELECTOR=""
 WAIT_MS="1500"
+# Set to "playwright", "puppeteer", or "chrome-mcp"
+backend="playwright"
 
 while (($#)); do
   case "$1" in
-    --url) URL="$2"; shift 2 ;;
-    --out) OUT="$2"; shift 2 ;;
+    --url)     URL="$2";     shift 2 ;;
+    --out)     OUT="$2";     shift 2 ;;
     --selector) SELECTOR="$2"; shift 2 ;;
-    --wait) WAIT_MS="$2"; shift 2 ;;
-    *) agc::die "unknown flag: $1" ;;
+    --wait)    WAIT_MS="$2"; shift 2 ;;
+    *)         agc::die "unknown flag: $1" ;;
   esac
 done
 
@@ -40,48 +44,102 @@ done
 
 mkdir -p "$(dirname "$OUT")"
 
-agc::log "browser-verify url=${URL} out=${OUT} viewport=${AGC_BROWSER_VIEWPORT} headless=${AGC_BROWSER_HEADLESS}"
-
-# --------------------------------------------------------------------------
-# Implementation hook — pick ONE backend and uncomment.
-# Left as stubs on purpose; user will plug in preferred tooling.
-# --------------------------------------------------------------------------
-
-# === Backend A: Playwright CLI =============================================
-# agc::require npx
-# VIEWPORT_W="${AGC_BROWSER_VIEWPORT%x*}"
-# VIEWPORT_H="${AGC_BROWSER_VIEWPORT#*x}"
-# npx --yes playwright@latest screenshot \
-#   --viewport-size="${VIEWPORT_W},${VIEWPORT_H}" \
-#   --wait-for-timeout "$WAIT_MS" \
-#   ${SELECTOR:+--selector "$SELECTOR"} \
-#   "$URL" "$OUT"
-
-# === Backend B: Puppeteer via node =========================================
-# node "$(dirname "$0")/puppet.mjs" --url "$URL" --out "$OUT" \
-#   --selector "$SELECTOR" --wait "$WAIT_MS" \
-#   --viewport "$AGC_BROWSER_VIEWPORT" --headless "$AGC_BROWSER_HEADLESS"
-
-# === Backend C: chrome-devtools MCP ========================================
-# The agent should call the MCP tool directly rather than shelling out.
-# If routed through this script, print the hint so the agent knows to switch:
-# agc::log "use mcp__chrome_devtools__take_screenshot instead of this shell"
-
-# --------------------------------------------------------------------------
-# Fallback: stub mode
-# --------------------------------------------------------------------------
-if [[ ! -s "$OUT" ]]; then
+VIEWPORT_W="${AGC_BROWSER_VIEWPORT%x*}"
+VIEWPORT_H="${AGC_BROWSER_VIEWPORT#*x}"
+
+agc::log "browser-verify backend=${backend} url=${URL} out=${OUT} viewport=${AGC_BROWSER_VIEWPORT} headless=${AGC_BROWSER_HEADLESS}"
+
+# ---------------------------------------------------------------------------
+# Backend A: Playwright CLI  (default)
+# ---------------------------------------------------------------------------
+playwright_backend() {
+  agc::require npx
+  local extra=()
+  [[ -n "$SELECTOR" ]] && extra+=(--selector "$SELECTOR")
+
+  if npx --yes playwright@latest screenshot \
+    --viewport-size="${VIEWPORT_W},${VIEWPORT_H}" \
+    --wait-for-timeout "$WAIT_MS" \
+    "${extra[@]}" \
+    "$URL" "$OUT" 2>/dev/null; then
+    return 0
+  fi
+  return 1
+}
+
+# ---------------------------------------------------------------------------
+# Backend B: Puppeteer via node
+# ---------------------------------------------------------------------------
+puppeteer_backend() {
+  if [[ -f "$(dirname "$0")/puppet.mjs" ]]; then
+    node "$(dirname "$0")/puppet.mjs" \
+      --url "$URL" --out "$OUT" \
+      --selector "$SELECTOR" --wait "$WAIT_MS" \
+      --viewport "$AGC_BROWSER_VIEWPORT" \
+      --headless "$AGC_BROWSER_HEADLESS" 2>/dev/null
+    return $?
+  fi
+  return 1
+}
+
+# ---------------------------------------------------------------------------
+# Backend C: chrome-devtools MCP
+# ---------------------------------------------------------------------------
+chrome_mcp_backend() {
+  # The agent should call mcp__chrome_devtools__take_screenshot directly.
+  # If routed here, just log the hint — we can't invoke MCP from bash.
+  agc::log "chrome-mcp: use mcp__chrome_devtools__take_screenshot tool directly"
+  return 1
+}
+
+# ---------------------------------------------------------------------------
+# Run the selected backend; fall back to stub on failure
+# ---------------------------------------------------------------------------
+did_capture=false
+
+case "$backend" in
+  playwright)
+    if playwright_backend; then
+      did_capture=true
+      agc::log "screenshot captured via Playwright: $OUT"
+    else
+      agc::log "Playwright capture failed, falling back to stub"
+    fi
+    ;;
+  puppeteer)
+    if puppeteer_backend; then
+      did_capture=true
+      agc::log "screenshot captured via Puppeteer: $OUT"
+    else
+      agc::log "Puppeteer capture failed, falling back to stub"
+    fi
+    ;;
+  chrome-mcp)
+    if chrome_mcp_backend; then
+      did_capture=true
+    else
+      agc::log "chrome-devtools MCP unavailable, falling back to stub"
+    fi
+    ;;
+  *)
+    agc::die "unknown backend: $backend (expected playwright|puppeteer|chrome-mcp)"
+    ;;
+esac
+
+# ---------------------------------------------------------------------------
+# Stub fallback
+# ---------------------------------------------------------------------------
+if [[ "$did_capture" != "true" ]]; then
   cat > "${OUT%.png}.stub.txt" <<EOF
-# TODO(auto-gh): browser-verify backend not wired up yet.
-# Requested: $URL
+# browser-verify backend not captured for: $URL
+# Backend: $backend
 # Would write screenshot to: $OUT
 # Selector: ${SELECTOR:-<full page>}
 # Wait: ${WAIT_MS}ms
 # Viewport: ${AGC_BROWSER_VIEWPORT}
 EOF
-  agc::log "stub: browser backend not configured (see ${OUT%.png}.stub.txt)"
-  printf 'SCREENSHOT=%s\n' "${OUT%.png}.stub.txt"
-  exit 0
+  agc::log "stub: browser capture unavailable (see ${OUT%.png}.stub.txt)"
 fi
 
 printf 'SCREENSHOT=%s\n' "$OUT"
+exit 0
\ No newline at end of file
diff --git a/skills/auto-github-contributor/scripts/dev-loop-check.sh b/skills/auto-github-contributor/scripts/dev-loop-check.sh
index 00ed63a..2933424 100755
--- a/skills/auto-github-contributor/scripts/dev-loop-check.sh
+++ b/skills/auto-github-contributor/scripts/dev-loop-check.sh
@@ -1,9 +1,11 @@
 #!/usr/bin/env bash
+
 # Phase-aware dev-loop checker. Invoked from the repo workdir.
+#
 # Phases:
-#   --phase red    → expect the newest test(s) to FAIL (proves coverage gap)
-#   --phase green  → install + lint + typecheck + test (must all pass)
-#   --phase final  → clean install + lint + typecheck + full tests + build
+# --phase red  → expect the newest test(s) to FAIL (proves coverage gap)
+# --phase green → install + lint + typecheck + test (must all pass)
+# --phase final → clean install + lint + typecheck + full tests + build
 #
 # Usage:
 #   dev-loop-check.sh --phase green [--workdir <dir>]
@@ -42,20 +44,20 @@ run() {
 
 # Pick the right package manager command based on lockfile.
 detect_install_cmd() {
-  if [[ -f pnpm-lock.yaml ]]; then echo "pnpm install --frozen-lockfile";
-  elif [[ -f yarn.lock ]]; then echo "yarn install --frozen-lockfile";
-  elif [[ -f package-lock.json ]]; then echo "npm ci";
-  elif [[ -f bun.lockb ]]; then echo "bun install --frozen-lockfile";
-  else echo "${AGC_INSTALL_CMD}";
+  if [[ -f pnpm-lock.yaml ]];     then echo "pnpm install --frozen-lockfile"
+  elif [[ -f yarn.lock ]];        then echo "yarn install --frozen-lockfile"
+  elif [[ -f package-lock.json ]]; then echo "npm ci"
+  elif [[ -f bun.lockb ]];        then echo "bun install --frozen-lockfile"
+  else                                 echo "${AGC_INSTALL_CMD}"
   fi
 }
 
 # Pick the JS package manager runner based on lockfile.
 detect_runner() {
-  if [[ -f pnpm-lock.yaml ]]; then echo "pnpm run";
-  elif [[ -f yarn.lock ]]; then echo "yarn";
-  elif [[ -f bun.lockb ]]; then echo "bun run";
-  else echo "npm run";
+  if [[ -f pnpm-lock.yaml ]]; then echo "pnpm run"
+  elif [[ -f yarn.lock ]];    then echo "yarn"
+  elif [[ -f bun.lockb ]];    then echo "bun run"
+  else                             echo "npm run"
   fi
 }
 
@@ -76,6 +78,7 @@ TEST_CMD="$(detect_script test "$AGC_TEST_CMD")"
 BUILD_CMD="$(detect_script build "$AGC_BUILD_CMD")"
 
 case "$PHASE" in
+
   red)
     # Red: run only the test suite. Expect non-zero (failing new test) — we
     # invert the exit code so a "correctly failing" red phase reports success.
@@ -90,24 +93,38 @@ case "$PHASE" in
     ;;
 
   green)
-    run "install" bash -lc "$INSTALL_CMD"
-    run "lint" bash -lc "$LINT_CMD"
-    run "typecheck" bash -lc "$TYPECHECK_CMD"
-    run "test" bash -lc "$TEST_CMD"
+    run "install"     bash -lc "$INSTALL_CMD"
+    run "lint"        bash -lc "$LINT_CMD"
+    run "typecheck"   bash -lc "$TYPECHECK_CMD"
+    run "test"        bash -lc "$TEST_CMD"
     ;;
 
   final)
-    run "install (clean)" bash -lc "$INSTALL_CMD"
-    run "lint" bash -lc "$LINT_CMD"
-    run "typecheck" bash -lc "$TYPECHECK_CMD"
-    run "test" bash -lc "$TEST_CMD"
-    run "build" bash -lc "$BUILD_CMD"
-    # TODO(auto-gh): repo-specific `verify` / e2e hook. Wire a shell command here
-    # once the target repo ships a stable e2e target (e.g. `pnpm test:e2e`).
+    run "install (clean)"  bash -lc "$INSTALL_CMD"
+    run "lint"             bash -lc "$LINT_CMD"
+    run "typecheck"        bash -lc "$TYPECHECK_CMD"
+    run "test"             bash -lc "$TEST_CMD"
+    run "build"            bash -lc "$BUILD_CMD"
+
+    # Wired e2e / verify hook if the target repo defines one.
+    # Looks for: test:e2e, test:e2e:ci, or a top-level `verify` script.
+    if [[ -f package.json ]]; then
+      E2E_SCRIPT="$(jq -r '
+        .scripts["test:e2e"] // .scripts["test:e2e:ci"] // .scripts.verify // empty
+      ' package.json 2>/dev/null || echo '')"
+      if [[ -n "$E2E_SCRIPT" ]] && [[ "$E2E_SCRIPT" != "null" ]]; then
+        RUNNER="$(detect_runner)"
+        run "e2e" bash -lc "$RUNNER $E2E_SCRIPT"
+      else
+        agc::log "e2e: no test:e2e / verify script found, skipping"
+      fi
+    fi
+
     agc::log "final verification complete"
     ;;
 
   *)
     agc::die "invalid phase: $PHASE (expected red|green|final)"
     ;;
-esac
+
+esac
\ No newline at end of file