Skip to content
View edwardjgriggs's full-sized avatar

Highlights

  • Pro

Block or report edwardjgriggs

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
edwardjgriggs/README.md

Edward Griggs

Security Operations & Detection Engineering | AI Security Researcher

Security+ certified analyst who defends a federal GovCon Microsoft 365 environment and builds the detections that catch attacks against it. I pair hands-on security operations (incident response, identity governance, NIST SP 800-171 compliance) with detection engineering in Microsoft Sentinel, and published AI security research on prompt injection and LLM defenses. I also build AI tools that keep a human in the loop.

Currently building out my detection portfolio and studying for SC-200.

Detection engineering portfolio

Four repositories that cover the full lifecycle of a detection, from written rule to a closed case.

  • detection-engineering Tuned KQL detections for Microsoft Sentinel and Defender, each mapped to MITRE ATT&CK with threat context, false-positive notes, and validation steps.
  • soc-detection-lab The self-built Sentinel and Wazuh lab where every detection is validated against real telemetry before I trust it.
  • detection-as-code Version-controlled Sigma rules with a CI pipeline that validates and converts them to KQL on every commit.
  • soc-casebook AI first-pass triage with a human in the loop. The model recommends, a named analyst approves or rejects every containment action, and every alert ends in a written case file.

Write the detection, prove it in the lab, automate the delivery, own the decision. That is the loop.

AI security research

Hijacking the Prompt: A Survey of Prompt Injection Attacks, Detection, and Defense in LLMs Published through ODU Digital Commons. Includes a Python/Flask proof-of-concept classifier matching 13 injection signatures to a six-category taxonomy, validated by 83 passing pytest tests. Survey, interactive demo, and conference presentation live in prompt-injection-research.

Things I've built

Security and compliance

  • widedoor Open source accessibility checker for websites and PDFs against Section 508 and WCAG 2.1 AA. Plain-language findings, and ACR output in OpenACR and the VPAT 2.5 Section 508 layout.
  • ttb-label-checker Batch alcohol label review. GPT-4o vision reads the label, a deterministic TypeScript rules engine judges every TTB check. The AI reads; code judges.
  • vault-7 Team Lead and Lead Developer for a four-person, browser-based security training game with mission-style scenarios on phishing, passwords, and defensive decisions. Winning project in the COVA CCI Design Cohort.

Human-in-the-loop AI

  • flight-sentry Mission assurance for spacecraft telemetry using trusted context and IBM Granite, with a human making the final call.
  • selvedge An AI that defends your voice instead of averaging it away. Built for the IBM SkillsBuild AI Builders Challenge, July 2026.

AI developer tools

  • prompt-lathe Shapes a raw prompt to a specific model, environment, and task. Desktop, CLI, and web share one local rules engine, with no telemetry.
  • skill-forge Create, validate, and publish agent skills for Claude, Grok, Cursor, and others.
  • SayLine Private, offline voice dictation for Windows. Hold a shortcut, speak, and local Whisper transcription types into any app. No account, no API key.

Apps and games

  • PromptLab iOS app that teaches better prompting through guided questions, lessons, and a built-in coach.
  • PacketGlide One-tap arcade game for iPhone, built with Godot. Free on the App Store.

What I work on

Security operations and compliance Incident and phishing response, Microsoft Defender, Entra ID, Intune, and Purview DLP across a 50+ user GovCon environment. Primary author behind a perfect 110/110 NIST SP 800-171 SPRS score, with the organization at CMMC Level 1 today and actively preparing for Level 2. Raised Microsoft Secure Score from 54.5% to 83.6% across 30+ endpoints.

Security automation PowerShell, Python, Power Automate, and the Microsoft Graph API to automate identity lifecycle and IT operations, recovering 1,200+ staff-hours per year.

AI governance Authoring enterprise AI use policies and advising on safe, accountable AI adoption.

Core skills

Security Operations · Detection Engineering · Microsoft Sentinel · KQL · Sigma · MITRE ATT&CK · Wazuh · Detection as Code · Incident Response · Microsoft Defender · Entra ID · Intune · Purview DLP · Conditional Access · NIST SP 800-171 · CMMC (L1 today, L2 in prep) · Section 508 / WCAG 2.1 · PowerShell · Python · TypeScript · C# / .NET · Swift · Microsoft Graph API · Prompt Injection Defense · LLM Security · Human-in-the-Loop AI

Certifications

  • CompTIA Security+ (SY0-701), 2026
  • Google Cybersecurity Professional Certificate (V2), 2025
  • Google AI Professional Certificate, 2025
  • IBM Generative AI Essentials for Cybersecurity, 2025
  • Microsoft SC-200: Security Operations Analyst (in progress)

Connect

LinkedIn Website Email

Pinned Loading

  1. prompt-injection-research prompt-injection-research Public

    Prompt injection security research — literature survey, interactive demo, and conference presentation for Virginia's Commonwealth Cyber Initiative

    Python

  2. detection-engineering detection-engineering Public

    A catalog of documented, tuned, ATT&CK-mapped KQL detections for Microsoft Sentinel and Defender.

  3. soc-casebook soc-casebook Public

    AI first-pass SOC triage with a human in the loop. Every alert ends in a case file.

    Python

  4. vault-7 vault-7 Public

    Vault 7 — interactive cybersecurity training game built for Virginia's Commonwealth Cyber Initiative (CCI) Design Cohort, Team 7.

    HTML

  5. widedoor widedoor Public

    Open source accessibility checker for websites and PDFs (Section 508 and WCAG 2.1 AA)

    TypeScript