Pre-launch — Gluecron is in final validation. Public signups and git hosting for non-owner users open after launch review.
CodeIssuesDiscussionsWikiPull RequestsProjectsCommitsActionsReleasesContributorsPulse● GatesSecuritySettingsDeploymentsPipelineInsightsAgents✨ Explain✨ Ask AI✨ Workspace✨ Spec✨ Tests▓ Debt Map✨ NL Search🏛 Archaeology
claude/adoring-hopper-5x74bqclaude/affectionate-feynman-ykrf1hclaude/architecture-audit-design-wxprenclaude/build-status-update-3MXsfclaude/charming-meitner-mllb5rclaude/compare-gate-gluecron-s4mFQclaude/confident-faraday-tikcwbclaude/continue-work-XMTlIclaude/crontech-gluecron-deploy-7MIECclaude/crontech-platform-setup-SeKfwclaude/design-2026claude/ecstatic-ptolemy-jMdigclaude/enhance-github-integration-QNHdGclaude/fix-aa-loop-issue-PonMQclaude/fix-actions-and-processclaude/fix-desktop-errors-XqoW8claude/fix-red-workflowsclaude/fix-website-access-6FKJNclaude/gatetest-integration-hardeningclaude/github-audit-improvements-bDFr9claude/gluecron-launch-status-FoMRlclaude/hopeful-lamport-olfCTclaude/issue-to-pr-and-protectionsclaude/jolly-heisenberg-2sg1Qclaude/launch-preparation-QmTb6claude/new-session-xk1l7claude/plan-platform-architecture-kkN4yclaude/platform-analysis-roadmap-1nUGLclaude/platform-launch-assessment-8dWV8claude/polish-platform-release-AeDrUclaude/resume-previous-work-KzyLwclaude/review-crontech-handoff-qYEVqclaude/review-project-completeness-lHhS2claude/review-readme-docs-ulqPKclaude/serene-edison-rj87weclaude/setup-multi-repo-dev-BCwNQclaude/ship-fixes-and-tests-Jvz1cclaude/site-audit-competitive-pctlwgclaude/site-migration-vercel-XstpKclaude/standalone-product-repos-XHFTDcopilot/feat-smart-empty-states-keyboard-first-enhancementcopilot/feat-smart-morning-digest-review-context-restorecopilot/fix-and-process-workflowscopilot/update-ai-powered-code-reviewfeat/debt-mapfeat/push-policy-codeowners-hardeningfeat/smart-digest-contextfeat/stage-impactfeat/t1-secret-migrationfeat/u-polishfeat/w-self-hostfeat/w2-claude-configfix/agent-journey-orphan-sweepgatetest/auto-fix-1776586424172gatetest/auto-fix-1776586534814gatetest/auto-fix-1776590685143gatetest/auto-fix-1776590808199mainops/redeploy-retriggerstyle/dxt-cta-themeworktree-agent-a3377aad30d55da26worktree-agent-a7ef607b7ee1d6c74
vultr-deploy.yml7.6 KB · 183 lines
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
name: Vultr Deploy (gluecron.com)

# Triggered on every push to main and on manual dispatch.
# Steps:
#   1. Capture the current SHA on the box (for rollback)
#   2. SSH in, git pull, run deploy-crontech.sh (which restarts systemd + reloads Caddy)
#   3. Smoke-test https://gluecron.com/healthz with retries
#   4. On smoke failure: roll back to the previous SHA and restart
#   5. On any failure: have Claude read the last 100 journal lines and post a one-paragraph
#      root-cause analysis to the workflow summary (and optionally a webhook)
#
# Secrets required:
#   VULTR_HOST      — public IP/hostname of the box (e.g. 45.76.171.37)
#   VULTR_USER      — ssh user (e.g. root)
#   VULTR_SSH_KEY   — private deploy key (PEM/OpenSSH format)
#
# Optional:
#   ANTHROPIC_API_KEY      — enables AI failure-diagnosis step
#   DEPLOY_WEBHOOK_URL     — POSTed with JSON deploy status (Slack/Discord/anything)

on:
  push:
    branches: [main]
  workflow_dispatch: {}

concurrency:
  group: vultr-deploy
  cancel-in-progress: false

jobs:
  deploy:
    name: Deploy gluecron.com
    runs-on: ubuntu-latest
    timeout-minutes: 12
    steps:
      - uses: actions/checkout@v4

      # ─── 1. Capture pre-deploy SHA so we can rollback ───────────────────
      - name: Capture pre-deploy SHA
        id: prev
        uses: appleboy/ssh-action@v1.2.0
        with:
          host: ${{ secrets.VULTR_HOST }}
          username: ${{ secrets.VULTR_USER }}
          key: ${{ secrets.VULTR_SSH_KEY }}
          script_stop: true
          script: |
            cd /opt/gluecron
            sha=$(git rev-parse HEAD)
            echo "Previous SHA: $sha"
            echo "$sha" > /tmp/gluecron_prev_sha
            cat /tmp/gluecron_prev_sha

      # ─── 2. Deploy: pull main, run the production script ────────────────
      - name: Deploy
        id: deploy
        uses: appleboy/ssh-action@v1.2.0
        with:
          host: ${{ secrets.VULTR_HOST }}
          username: ${{ secrets.VULTR_USER }}
          key: ${{ secrets.VULTR_SSH_KEY }}
          command_timeout: 8m
          script_stop: true
          script: |
            set -euo pipefail
            cd /opt/gluecron
            git fetch --prune origin main
            git reset --hard origin/main
            new_sha=$(git rev-parse HEAD)
            echo "Deploying SHA: $new_sha"
            bash scripts/deploy-crontech.sh

      # ─── 3. Smoke-test the live URL ─────────────────────────────────────
      - name: Smoke test (https://gluecron.com/healthz)
        id: smoke
        run: |
          set +e
          for i in 1 2 3 4 5 6 7 8; do
            code=$(curl -s -o /dev/null -w "%{http_code}" https://gluecron.com/healthz)
            echo "Attempt $i: /healthz → $code"
            if [ "$code" = "200" ]; then
              echo "smoke_status=ok" >> $GITHUB_OUTPUT
              echo "✓ gluecron.com is live and healthy."
              exit 0
            fi
            sleep 8
          done
          echo "smoke_status=fail" >> $GITHUB_OUTPUT
          echo "✗ /healthz did not return 200 after 64s. Triggering rollback."
          exit 1

      # ─── 4. Auto-rollback on smoke failure ──────────────────────────────
      - name: Rollback on failure
        if: failure() && steps.smoke.conclusion == 'failure'
        uses: appleboy/ssh-action@v1.2.0
        with:
          host: ${{ secrets.VULTR_HOST }}
          username: ${{ secrets.VULTR_USER }}
          key: ${{ secrets.VULTR_SSH_KEY }}
          script_stop: true
          script: |
            set -euo pipefail
            cd /opt/gluecron
            prev=$(cat /tmp/gluecron_prev_sha)
            echo "Rolling back to $prev"
            git reset --hard "$prev"
            systemctl restart gluecron
            sleep 5
            systemctl status gluecron --no-pager | head -20

      # ─── 5. AI failure diagnosis (only on failure, only if ANTHROPIC set) ─
      - name: Capture failure context
        if: failure()
        id: ctx
        uses: appleboy/ssh-action@v1.2.0
        with:
          host: ${{ secrets.VULTR_HOST }}
          username: ${{ secrets.VULTR_USER }}
          key: ${{ secrets.VULTR_SSH_KEY }}
          script_stop: false
          script: |
            echo "===== systemd status ====="
            systemctl status gluecron --no-pager 2>&1 | head -30 || true
            echo "===== last 100 journal lines ====="
            journalctl -u gluecron -n 100 --no-pager 2>&1 || true
            echo "===== caddy reload check ====="
            caddy validate --config /etc/caddy/Caddyfile 2>&1 || true
            echo "===== /healthz from box ====="
            curl -s -w "HTTP %{http_code}\n" http://localhost:3000/healthz 2>&1 || true

      - name: AI root-cause analysis (Claude)
        if: failure() && env.ANTHROPIC_API_KEY != ''
        env:
          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
          DEPLOY_LOGS: ${{ steps.ctx.outputs.stdout }}
          COMMIT_SHA: ${{ github.sha }}
          COMMIT_MSG: ${{ github.event.head_commit.message }}
        run: |
          set +e
          # Build the prompt
          cat > /tmp/prompt.json <<EOF
          {
            "model": "claude-haiku-4-5-20251001",
            "max_tokens": 600,
            "system": "You are a senior SRE diagnosing a failed deploy. Read the systemd status, journal, and curl output. In 1 short paragraph (under 100 words), identify the most likely root cause and the single fastest fix. Be direct, no preamble.",
            "messages": [{
              "role": "user",
              "content": "Commit: $COMMIT_SHA\nMessage: $COMMIT_MSG\n\nDeploy logs:\n$DEPLOY_LOGS"
            }]
          }
          EOF
          response=$(curl -s https://api.anthropic.com/v1/messages \
            -H "x-api-key: $ANTHROPIC_API_KEY" \
            -H "anthropic-version: 2023-06-01" \
            -H "content-type: application/json" \
            --data @/tmp/prompt.json)
          analysis=$(echo "$response" | python3 -c "import sys,json; d=json.load(sys.stdin); print(d.get('content',[{}])[0].get('text','(no analysis)'))" 2>/dev/null)
          echo "## 🤖 AI Failure Analysis" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "$analysis" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "**Commit:** \`${COMMIT_SHA:0:7}\` — $COMMIT_MSG" >> $GITHUB_STEP_SUMMARY

      # ─── 6. Optional webhook notification ───────────────────────────────
      - name: Notify webhook
        if: always() && env.DEPLOY_WEBHOOK_URL != ''
        env:
          DEPLOY_WEBHOOK_URL: ${{ secrets.DEPLOY_WEBHOOK_URL }}
          STATUS: ${{ job.status }}
        run: |
          curl -s -X POST "$DEPLOY_WEBHOOK_URL" \
            -H "content-type: application/json" \
            --data "{\"status\":\"$STATUS\",\"target\":\"gluecron.com\",\"sha\":\"${{ github.sha }}\",\"run\":\"${{ github.run_id }}\"}" || true

      # ─── 7. Workflow summary on success ─────────────────────────────────
      - name: Success summary
        if: success()
        run: |
          echo "## ✅ Deploy succeeded" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          echo "- **Target:** https://gluecron.com" >> $GITHUB_STEP_SUMMARY
          echo "- **SHA:** \`${GITHUB_SHA:0:7}\`" >> $GITHUB_STEP_SUMMARY
          echo "- **Health:** /healthz → 200" >> $GITHUB_STEP_SUMMARY