alcyone-labs/aquaria-cloudflare-ops
skills/aquaria-cloudflare-ops/SKILL.md
Operate and debug Aquaria Cloudflare Worker + Workflows deployment. Deploy safely run health checks inspect traces restart stuck instances apply fixes.
$ install --global
skillsauthnpx skillsauth add alcyone-labs/agent-skills aquaria-cloudflare-opsInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 2, 2026, 4:47 AM62.3s9 files scanned
SKILL.md
- name:
- aquaria-cloudflare-ops
- description:
- Operate and debug Aquaria Cloudflare Worker + Workflows deployment. Deploy safely run health checks inspect traces restart stuck instances apply fixes.
Aquaria Cloudflare Ops (rss-feed-processor)
When to apply
- Deploy
workflows/packages/rss-feed-processorWorker - Check production health daily checks workflow failures
- Debug Cloudflare Workflows
item-process-workflowfeed-poll-workflowqa-analysis-workflowdaily-summary - Mitigate incidents kill switch bulk terminate restart stuck instances
Inputs / prerequisites
Provide one of:
- Project path (preferred):
--project workflows/packages/rss-feed-processor - Explicit URL:
--url https://<worker>.<subdomain>.workers.dev
Credentials:
- Admin key for
/api/admin/*and/api/traces/*:--admin-keyflag or envAQUARIA_ADMIN_KEY/ADMIN_KEY
- Cloudflare API token for deep inspection and restart fallback:
- env
CLOUDFLARE_API_TOKEN
- env
Rules
- Always start narrow then widen failure window (
--since 30mbefore24h) - Never print
.env*contents or secrets in terminal output - Quote operational lessons from
~/.factory/sessions/*.jsonlnot inline secrets - Use
--step-output=falsefor Wrangler to avoid dumping huge outputs - Enable kill switch before investigating incidents
Workflow decision tree
Request: deploy rss-feed-processor
A. Pre-deploy: test → build → kill-switch enable (if incident)
B. Deploy: wrangler deploy → capture version ID
C. Post-deploy: verify → daily-check 30m → trace list --status failure
D. If failures: trace show <instanceId> → diagnose → fix → redeploy
Request: check prod health
A. Start narrow: daily-check --since 30m
B. If issues: widen to 2h → trace list --status failure
C. If failures found: trace show <instanceId> → inspect step output
Request: debug stuck workflow
A. List: workflow list --workflow <name> --status errored
B. Inspect: trace show <instanceId> OR wrangler describe
C. Action: restart <instanceId> OR terminate <instanceId>
Request: incident response
A. Immediate (0-5m): kill-switch enable → verify → daily-check 1h
B. Diagnose (5-15m): trace list --status failure → trace show <instanceId>
C. Fix → deploy with kill-switch ON
D. Recovery: post-deploy checks → kill-switch disable only after stable
Examples
Input: Deploy rss-feed-processor Output:
pnpm -C workflows/packages/rss-feed-processor test -- --run --timeout 60000
pnpm -C workflows/packages/rss-feed-processor build
pnpm -C workflows/packages/rss-feed-processor run deploy
pnpm -s aquaria cloudflare --project workflows/packages/rss-feed-processor verify
pnpm -s aquaria cloudflare --project workflows/packages/rss-feed-processor daily-check --since 30m
Input: Production showing failures Output:
pnpm -s aquaria cloudflare --project workflows/packages/rss-feed-processor daily-check --since 30m
pnpm -s aquaria cloudflare --project workflows/packages/rss-feed-processor trace list --workflow item-process --since 30m --status failure
# Pick instanceId from output
pnpm -s aquaria cloudflare --project workflows/packages/rss-feed-processor trace show --instance-id <instanceId>
Input: SQLITE_TOOBIG error in traces Output:
# Confirm failing step
pnpm -s aquaria cloudflare --project workflows/packages/rss-feed-processor trace show --instance-id <instanceId>
# Or use Wrangler for full detail
pnpm -C workflows/packages/rss-feed-processor exec wrangler workflows instances describe item-process-workflow <instanceId> --step-output=false --truncate-output-limit 2000
# Fix: apply trace filtering to exclude large fields
# Deploy and watch
traces list --status failure --since 30m
Related Skills
alcyone-labs/skill-forge
development
VerifiedTrustedCommunity
Builds precise production-ready custom Agent Skills following AgentSkills.io guidelines. Use when user requests to create, refine or package Skills
1SKILL.mdUpdated Apr 2, 2026
alcyone-labs/simple-logger-usage
development
VerifiedTrustedCommunity
Best practices for using simple-logger in TypeScript applications
1SKILL.mdUpdated Apr 2, 2026
alcyone-labs/sauve-jazz-extension
tools
VerifiedTrustedCommunity
Implement Sauve-specific Jazz extension behavior on top of worker-authority architecture and typed popup-worker protocol
1SKILL.mdUpdated Apr 2, 2026
alcyone-labs/playwright-chrome-extension-testing
tools
VerifiedTrustedCommunity
Test Chrome extensions (Manifest V3) headlessly with Playwright — load unpacked extensions, extract the extension ID, test popups/content scripts/background, run reliably in CI, and use the correct headless mode that actually supports extensions.
1SKILL.mdUpdated Apr 2, 2026