Skip to content
View ShalvaLekishvili's full-sized avatar
🤒
Out sick
🤒
Out sick

Block or report ShalvaLekishvili

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. Markdown is supported. 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
ShalvaLekishvili/README.md
Python C# Java JavaScript C++ icon React Django GraphQL NumPy Pandas LangGraph Jupyter Keras OpenAI LangChain Hugging Face OpenCV Matplotlib MySQL Docker n8n Streamlit Canva

Profile View

Executive Brief

Security operations with engineering depth

I am a SOC Analyst (L2) focused on detection engineering, incident investigation, threat hunting, network forensics, security automation, and authorized security validation.

My portfolio is built around one rule:

A security finding is only useful when it produces evidence, context, scope, a decision, and a stronger control.

I want a reviewer to be able to inspect the engineering behind the claim—not just read that I know a tool.

Core operating domains

SOC Operations · Detection Engineering · DFIR · Threat Hunting · Network Forensics · Security Automation

Proof stack

Signal Public evidence
Curated detections 12
PacketScope tests 48
Documented credentials 14
Flagship platforms 2
Primary engineering language Python
Portfolio philosophy Evidence-first

Live Portfolio Intelligence

Live GitHub portfolio intelligence for SentinelForge and PacketScope

The panel above is generated by this repository, not by a third-party statistics service. A scheduled GitHub Actions workflow queries the GitHub REST API, verifies selected public claims against the flagship project READMEs, and regenerates local SVG assets plus a machine-readable manifest.

Machine-readable evidence: assets/generated/portfolio-manifest.json

The generator has an explicit baseline fallback. If GitHub API collection fails, the output is marked as baseline-backed rather than silently pretending stale information is live.


Flagship Security Engineering

Defensive SOC investigation + detection engineering workbench

SentinelForge converts blue-team claims into inspectable engineering: telemetry parsers, a normalized event model, detection-as-code, process correlation, IOC extraction, ATT&CK context, analyst workflows, API/CLI surfaces, testing, and CI.

Public engineering evidence

  • Windows .evtx ingestion
  • Wazuh / Sysmon-friendly normalization
  • External YAML detection rules
  • 12 curated defensive detections
  • Severity + confidence-aware risk scoring
  • IOC extraction for common indicator classes
  • PID / PPID process graph correlation
  • MITRE ATT&CK technique coverage
  • FastAPI + CLI + analyst workspace
  • Hardened container baseline
  • Tests, CI and dependency maintenance

Reviewer route: README · Architecture · Rules · Tests

Local-first network forensics + packet investigation workbench

PacketScope transforms PCAP / PCAPNG evidence into a structured analyst investigation: hosts, flows, transactions, protocol metadata, findings, evidence slices, analyst state, reports, and investigation pivots.

Public engineering evidence

  • PCAP + PCAPNG ingestion
  • DNS / HTTP / TLS / ARP / DHCP / ICMP / NTP metadata
  • Host, flow and conversation modeling
  • Bounded TCP reconstruction
  • Behavioral detection leads
  • IOC and transaction context
  • Packet-level evidence slicing
  • Analyst verdicts, notes and tags
  • JSON + self-contained HTML reporting
  • FastAPI + CLI + Docker
  • Local-first privacy model
  • 48 automated tests

Reviewer route: README · Architecture · Detections · Tests

Evidence Registry

Capability SentinelForge PacketScope What it demonstrates
Detection engineering Behavior translated into repeatable analyst logic
Windows / endpoint telemetry Event normalization, process relationships, command context
Network telemetry Protocol-aware traffic and conversation reconstruction
IOC extraction Investigation pivots and contextualization
Correlation Relationships rather than isolated raw events
ATT&CK context Technique-aware analyst framing
Analyst workflow Findings connected to review and evidence
API / CLI engineering Automation and integration surfaces
Automated testing Reproducible engineering rather than presentation-only claims
Defensive boundaries Explicit limitations and intended use

Business & Operational Impact

Technical work Security question Operational output Business value
Alert triage Is this signal meaningful? Evidence-backed disposition Reduces analyst noise and missed escalation
Detection engineering What behavior must surface consistently? Tested rule logic + ATT&CK context Improves repeatability and visibility
Endpoint investigation What executed and what changed? Process tree, timeline, persistence evidence Improves incident scope and containment
Network forensics What crossed the wire? Conversations, protocol metadata, packet evidence Clarifies communication and data movement
Threat hunting What may exist without an alert? Hypothesis-driven pivots and findings Exposes detection gaps
Authorized validation Does the control resist expected abuse? Reproducible finding + retest Converts assumptions into verified assurance
Automation What repetitive work can become deterministic? Scripts, APIs, CLI workflows, reports Improves speed and consistency

Operational Deliverables

Investigate

  • Evidence triage
  • Process-tree reconstruction
  • Command-line review
  • Persistence analysis
  • Timeline building
  • IOC extraction
  • Scope determination

Detect

  • Detection-as-code
  • Rule authoring
  • ATT&CK mapping
  • False-positive analysis
  • Detection testing
  • Gap identification
  • Explainable findings

Communicate

  • Incident summaries
  • Severity rationale
  • Escalation context
  • Remediation guidance
  • Analyst notes
  • Evidence preservation
  • Retest documentation

Security Operating System

Security operating system from telemetry to control improvement
TELEMETRY
   ↓
NORMALIZE + CORRELATE
   ↓
DETECT + INVESTIGATE
   ↓
ESTABLISH SCOPE + PRIORITY
   ↓
ESCALATE / CONTAIN / CLOSE
   ↓
TUNE DETECTION + HARDEN CONTROL + DOCUMENT + RETEST

The operating goal is not “generate more alerts.” It is reduce uncertainty until a defensible next action is clear.


Credentials & Professional Development

Fourteen cybersecurity credentials and technical training programs

The credentials stay fully visible. They support the engineering evidence above; they do not replace it.

TryHackMe — 10 credentials

Credential Completed Credential ID Domain
Advent of Cyber 2025 Dec 2025 THM-EGTUQF4CAI Multi-domain Security
CTI – CISA JCDC Triage, Fusion and Analysis Nov 2025 THM-JVR3U2HPKS Cyber Threat Intelligence
Security Engineer Sep 2025 THM-2MWEFSMO18 Security Engineering
SOC Level 1 Sep 2025 THM-YHFGKSGY28 Security Operations
SOC Level 2 Sep 2025 THM-TH8U8ODRGS Advanced SOC Operations
Offensive Pentesting Sep 2025 THM-O5YB6D7QMZ Security Validation
Web Fundamentals Sep 2025 THM-Y05LJ9JOTY Web Security
CompTIA PenTest+ Aligned Learning Path Sep 2025 THM-JSEWACT2EY Penetration Testing
Jr Penetration Tester Sep 2025 THM-DFCKXEC7AT Penetration Testing
Red Teaming Sep 2025 THM-HEX0VWNFTS Adversary Simulation

Additional technical training — 4 programs

Course / Certificate Provider Completed Domain
CompTIA PenTest+ Ethical Hacking Training Udemy Sep 2024 Penetration Testing
SuperMap GIS Solutions for Telecommunications SuperMap GIS Nov 2021 GIS / Telecommunications
Fundamentals of Computer Hacking infySEC Jun 2021 Cybersecurity Fundamentals
Cyber Security and Ethical Hacking Georgian Project Andromeda Oct 2020 Ethical Hacking

Technical Capability Architecture

Security Operations

SIEM
Wazuh
Windows Events
Sysmon
FortiGate
Triage
Escalation

Detection Engineering

Detection-as-Code
MITRE ATT&CK
Rule Testing
IOC Logic
Correlation
Risk Scoring
Behavior Analytics

DFIR & Hunting

Timeline Analysis
Process Trees
PCAP
Network Forensics
Threat Hunting
Evidence Handling
Protocol Analysis

Engineering

Python
PowerShell
FastAPI
CLI
Docker
Git
GitHub Actions


Engineering Principles

01 · Explainability

A detection should show why it matched, what evidence supports it, and what the analyst should verify next.

02 · Evidence Integrity

Conclusions should remain traceable to source events, packets, timelines, or reproducible validation steps.

03 · Testable Security

Rules, parsers, detections, and automation should be tested rather than presented as unsupported claims.

04 · Privacy by Default

Public demonstrations should prefer synthetic evidence, local processing, redaction, and bounded collection.

05 · Defensive Scope

Security tooling should support investigation, detection, evidence handling, remediation, and authorized validation.

06 · Operational Clarity

Every useful output should lead toward investigate, escalate, contain, tune, harden, document, retest, or close.

The Profile Itself Is Engineered

This repository is intentionally designed like a small production system.

config/portfolio.json
        │
        ├── identity + credentials
        └── flagship claim definitions
                │
                ▼
      scripts/profile_engine.py
                │
     ┌──────────┴──────────┐
     ▼                     ▼
GitHub REST API       README verification
     │                     │
     └──────────┬──────────┘
                ▼
     local dark/light SVGs
     portfolio-manifest.json
                │
                ▼
            README.md

Automation hardening

  • GitHub REST API version is explicitly pinned to 2026-03-10.
  • The workflow runs on a timezone-aware Asia/Tbilisi schedule and can also be triggered manually.
  • Workflow actions are pinned to full commit SHAs.
  • GITHUB_TOKEN receives only the repository-content permission needed to commit generated assets.
  • The workflow never uses pull_request_target.
  • The generator uses the Python standard library only—no runtime package-install supply chain.
  • Selected public claims are regex-verified against each flagship repository README.
  • Unit tests run before the generated profile is committed.
  • A validator checks local assets, SVG validity, credential count, duplicate IDs, and required portfolio sections.
  • Dependabot is configured for GitHub Actions maintenance.
  • CODEOWNERS covers workflow, script, and configuration changes.

Local developer commands

make profile    # generate from bundled verified baseline
make test       # run unit tests
make validate   # validate README/assets/credentials
make all        # run everything

See docs/GOD_MODE_ARCHITECTURE.md for the design.


Reviewer Routes

Reviewer routes for hiring manager, SOC lead and technical reviewer

Hiring manager — 60 seconds

Role → proof stack → flagship projects → credentials → business impact

SOC / Blue Team lead — 3 minutes

Detection coverage → analyst workflows → investigation model → operating system → deliverables

Technical reviewer — 5+ minutes

Repository architecture → security boundaries → tests → workflow hardening → generated evidence manifest


Current Development Vector

01  Detection engineering + SIEM rule quality
02  Advanced Windows / endpoint investigation
03  Cyber threat intelligence + IOC contextualization
04  Active Directory security
05  Network forensics + protocol analysis
06  Web application security validation
07  Python / PowerShell security automation
08  Incident reporting + operational playbooks

Portfolio Governance

This GitHub is maintained as a professional security engineering portfolio, not merely a repository collection.

  • Public claims should have inspectable evidence wherever possible.
  • Offensive techniques are presented only in authorized, defensive, educational, or lab context.
  • Production-sensitive data is excluded from public demonstrations.
  • Flagship repositories should document architecture, testing, limitations, security scope, and roadmap.
  • Synthetic evidence is preferred for demonstrations.
  • Self-ratings such as “95% expert” are deliberately avoided.
  • Stars and followers are not treated as proof of security engineering quality.
  • New portfolio claims should be added to the machine-readable configuration when they can be verified.

Contact

Open to security operations, detection engineering, threat hunting, and DFIR opportunities

Shalva Lekishvili Shalva Lekishvili Shalva Lekishvili Shalva Lekishvili
Security portfolio footer

Popular repositories Loading

  1. Android_Infection Android_Infection Public

    Shell 1

  2. Phisher Phisher Public

    HTML 1

  3. XI-WebGIS XI-WebGIS Public

    JavaScript 1

  4. SentinelForge SentinelForge Public

    Python 1 1

  5. PacketScope PacketScope Public

    Python 1

  6. ShalvaLekishvili ShalvaLekishvili Public

    Python 1