By Hisham Alrashdan Founder of InfoGleam | 20+ Years in Healthcare & Dental Systems Engineering
I love automation — but only the kind that earns trust.
In 2026, “AI automation” is everywhere, yet many implementations still feel like a demo. The AI makes a decision, pushes buttons, and you hope it didn’t do something unexpected at 2:00 AM.
This is why I’m building with the OpenClaw mindset. It’s not about replacing humans; it’s about upgrading workflows by adding judgment, consistency, and speed—while always leaving a “safety valve” for a human to step in.
This article introduces my latest project: a secure, fast comment-moderation pipeline for WordPress.
- Website: infogleam.com
- Project Repo: github.com/Hisham-InfoGleam/openclaw-wordpress-guardian-agent
The Vision: Automation with a Human Safety Valve
Most real-world workflows—especially in sensitive fields like healthcare or data management—follow a predictable pattern:
- The Event: A form is submitted, a comment is posted, or a support ticket is created.
- The Decision: We must approve, reject, or escalate.
- The Stakes: The decision must be fast, but mistakes are expensive.
That’s where Human-in-the-Loop (HITL) design shines. The system handles the obvious cases automatically, and for anything ambiguous, it kicks the decision to a human.
It’s not “Automation vs. Humans.” It’s Automation + Humans.
Why OpenClaw for Workflow Automation?
OpenClaw allows us to treat an AI model like a dependable automation component. It keeps the system deterministic and auditable. In practical terms, this means:
- Seamless Integration: Add AI to an existing workflow without a total rewrite.
- Flexibility: Swap moderation logic from “simple rules” to “AI-backed classification” with a single toggle.
- Guardrails: Implement strict timeouts, fallbacks, and human review steps.
The future won’t be one giant AI doing everything. It will be small, composable automations where AI is used where it’s strong and supervised where the stakes are real.
System Architecture: The “Guardian” Flow
The project acts as a “Guardian Agent” for WordPress comments. It moderates automatically when confidence is high and routes uncertain cases to Telegram for a one-click decision.
- Event: A visitor posts a comment on WordPress.
- Webhook: A custom WordPress plugin sends a signed HMAC payload to the Automation API.
- Analysis: The API validates the signature and decides:
- Approve/Block: Done automatically if the “brain” is confident.
- Escalate: If uncertain, it sends a message to the owner via Telegram.
- Human Action: The owner taps “Approve” or “Block” directly in Telegram.
- Resolution: The API updates the comment status via the WordPress REST API.
Technical Deep Dive: Trust & Security
1. The Webhook Trust Anchor
To prevent “endpoint spoofing,” the WordPress plugin signs the JSON body using HMAC SHA-256. The API verifies this signature before any processing occurs. This ensures that only your WordPress site can trigger the moderation flow.
2. The Moderation Brain (Heuristic + OpenClaw)
The API uses a provider-based approach. You can toggle between:
- Heuristic Mode: Fast, rule-based logic (regex, keyword lists).
- OpenClaw Mode: AI-backed moderation via an OpenAI-compatible gateway.
The system is designed for resilience. If the AI provider is down or times out, the system automatically falls back to heuristic rules or human review.
3. The Telegram Review Loop
When a decision is escalated, the owner receives a Telegram message with inline buttons. Security is maintained here using a secret_token configured in the Telegram Webhook, ensuring the API only listens to legitimate Telegram updates.
4. Writing back to WordPress
We use WordPress Application Passwords. These are stored as environment variables, keeping the credentials secure, revocable, and entirely separate from the codebase.
Roadmap: What’s Next?
This project is evolving in real increments:
- Phase 1 (Current): Stateless processing + Telegram human-in-the-loop.
- Phase 1.5: OpenClaw moderation provider + intelligent fallbacks.
- Phase 2: Persistence (Prisma/PostgreSQL), audit trails, and analytics dashboards.
If you’re building “automation you can trust,” I invite you to follow the journey on GitHub.
Inspiration & References
About the Author
Hisham Alrashdan is an expert in healthcare and dental software systems with over 20 years of experience. Through InfoGleam, he focuses on building secure, efficient, and human-centric automation tools that bridge the gap between complex data and actionable insights.
Leave a Reply