ci: make deploy webhook curl robust and observable (#484)

## Problem

The deploy step intermittently fails on CI with only:

```
Response:
HTTP Status: 000
Deployment request failed with status 000
```

`000` means curl never completed a connection — the Coolify box (which
rebuilds the image from git on deploy) has its API unreachable while a
prior build runs. But the step gave no way to confirm that: `-s`
silenced curl's error message and `|| true` swallowed the exit code.

## Changes

- **Surface the real cause** — `-sS` plus `--write-out
%{exitcode}`/`%{errormsg}`/timing, so failures print the curl exit code
(7 refused, 28 timeout, …) and message instead of a blind `000`.
- **Robust parsing** — body/metadata split on an explicit
`===CURL_META===` marker instead of fragile `tail`/`sed` line counting.
- **No false timeouts** — `--max-time` raised 120→300s so a
slow-but-alive box doesn't abort mid-trigger and report a phantom `000`.
- **Wider retry window** — 10 attempts with backoff capped at 60s (~8
min) instead of 5 with a 240s dead sleep, to ride out a multi-minute
rebuild.
- **Fail fast on 4xx** — a bad token/uuid is not retryable, so don't
waste the window on it.

## Note

This makes the trigger robust and observable; the structural fix (let
Coolify pull the prebuilt ghcr image instead of rebuilding on the box)
is out of scope here.

.github/workflows/ci.yml

@@ -588,33 +588,58 @@ jobs:
       - name: Trigger deployment
         if: steps.deploy_guard.outputs.should_deploy == 'true'
         run: |
-          max_attempts=5
+          # The Coolify box rebuilds the image from git on deploy, so its API
+          # (port 8000) can be unreachable for several minutes while a prior
+          # build runs. We just need the webhook to land once: retry on
+          # connection-level failures with a capped backoff over a wide window.
+          max_attempts=10
+          max_delay=60
           attempt=1
 
+          # Separator we control, so curl's --write-out fields can never be
+          # confused with the response body.
+          marker="===CURL_META==="
+
           while [ $attempt -le $max_attempts ]; do
             echo "Attempt $attempt of $max_attempts..."
 
-            response=$(curl -s -w "\n%{http_code}" \
+            curl_exit=0
+            response=$(curl -sS \
+              -w "${marker}\nhttp_code=%{http_code}\ncurl_exit=%{exitcode}\nerrormsg=%{errormsg}\ntiming=DNS:%{time_namelookup}s Connect:%{time_connect}s Total:%{time_total}s Remote:%{remote_ip}:%{remote_port}" \
               --connect-timeout 30 \
-              --max-time 120 \
+              --max-time 300 \
               -H "Authorization: Bearer ${{ secrets.DEPLOYMENT_TOKEN }}" \
-              "http://147.93.126.54:8000/api/v1/deploy?uuid=ww00sswosco8w80k08c0occ8&force=false") || true
+              "http://147.93.126.54:8000/api/v1/deploy?uuid=ww00sswosco8w80k08c0occ8&force=false") || curl_exit=$?
 
-            http_code=$(echo "$response" | tail -n1)
-            body=$(echo "$response" | sed '$d')
+            body="${response%%${marker}*}"
+            meta="${response#*${marker}}"
+            http_code=$(echo "$meta" | sed -n 's/^http_code=//p')
+            curl_code=$(echo "$meta" | sed -n 's/^curl_exit=//p')
+            errormsg=$(echo "$meta" | sed -n 's/^errormsg=//p')
+            timing=$(echo "$meta" | sed -n 's/^timing=//p')
 
-            echo "Response: $body"
-            echo "HTTP Status: ${http_code:-timeout}"
+            echo "Response body: $body"
+            echo "HTTP Status: ${http_code:-none}"
+            echo "curl exit: ${curl_code:-$curl_exit}"
+            echo "curl error: ${errormsg:-none}"
+            echo "Timing: $timing"
 
-            if [ -n "$http_code" ] && [ "$http_code" -ge 200 ] && [ "$http_code" -lt 300 ]; then
+            if [ "${http_code:-0}" -ge 200 ] && [ "${http_code:-0}" -lt 300 ]; then
               echo "Deployment triggered successfully!"
               exit 0
             fi
 
-            echo "Deployment request failed with status $http_code"
+            # 4xx are caller errors (bad token / uuid) — retrying won't help.
+            if [ "${http_code:-0}" -ge 400 ] && [ "${http_code:-0}" -lt 500 ]; then
+              echo "Got client error $http_code — not retryable. Failing fast."
+              exit 1
+            fi
+
+            echo "Deployment request failed (http=${http_code:-none} curl_exit=${curl_code:-$curl_exit}: ${errormsg:-none})"
 
             if [ $attempt -lt $max_attempts ]; then
-              delay=$((30 * (2 ** (attempt - 1))))
+              delay=$((15 * attempt))
+              [ $delay -gt $max_delay ] && delay=$max_delay
               echo "Retrying in ${delay}s..."
               sleep $delay
             fi