I'm a software developer specializing in Python, AI tooling, and security engineering. I contribute across a wide range of open-source projects — from AI evaluation frameworks and LLM providers to Kubernetes RBAC scanners, Android apps, and ML visualization tools. I care about correctness, security, and leaving every codebase better than I found it.
name : Maqbool Ahmed
location : Pakistan
role : Software Developer
languages : Python · TypeScript · JavaScript · Go · Kotlin
domains : AI/LLM · Security Engineering · MCP Tooling · DevOps · Android
contributions: 110 merged PRs across 18+ repositories
status : Open to collaboration ✓
| # | Repository | What I did | Tags |
|---|---|---|---|
#627 |
Fmarzochi/EGC | Encrypted .md state files at rest with AES-256-GCM |
  |
#625 |
Fmarzochi/EGC | HMAC integrity checks on state files — tamper detection at rest |  |
#618 |
Fmarzochi/EGC | Built egc replay — full session playback with timeline scrubbing |
  |
#724 |
Fmarzochi/EGC | Native DeepSeek LLM provider with temperature normalization & exception retagging |   |
#3191 |
bernstein | Signed, independently-replayable trajectory receipts for AI eval |   |
#3099 |
bernstein | Shipped bernstein-bench — runnable, reproducibility-gated evaluation harness |
 |
#105 |
kaaval/kaaval | RBAC combination-escalation predicates — escalate, bind, impersonate, privileged pod |  |
#136 |
evaltrust | Streamed large JSONL/CSV files line-by-line instead of full memory read |   |
#30 |
open-devops-agent | Unit tests for storage providers, webhooks, MCP registry — fixed silent skips |   |
#143 |
visualtorch | Added outline_width param to flow, lenet, and graph neural net views |
 |
#3913 |
Trail-Sense | Show distance, direction, and elevation diff on long-press map point |   |
#165 |
matryca-plumber | Exposed public API on SessionAliasRegistry — removed SLF001 violations |
  |
Contributions span AI evaluation, LLM providers, security engineering, Kubernetes RBAC, Android, ML visualization, MCP tooling, DevOps, frontend, and more.
Python ████████████████████░░ 92% evaltrust · visualtorch · matryca-plumber · Termstory
AI / LLM Tooling ██████████████████░░░░ 88% DeepSeek provider · bernstein · evaltrust · MCP
Security Engineering █████████████████░░░░░ 85% AES-256-GCM · HMAC · RBAC · audit logs · Kubernetes
TypeScript / JavaScript ████████████████░░░░░░ 82% EGC · ansvisor · Prism-platform · layerkit
Testing & Reliability ███████████████░░░░░░░ 80% open-devops-agent · matryca-plumber · evaltrust
Go / Kubernetes █████████████░░░░░░░░░ 72% kaaval RBAC scanner · cluster security rules
Android / Kotlin ████████████░░░░░░░░░░ 68% Trail-Sense · UI · maps · battery detection
focus = {
"🤖 AI & LLM Tooling" : "Native LLM providers, eval harnesses, reproducible trajectory receipts (EGC, bernstein)",
"🔒 Security Engineering" : "AES-256-GCM encryption, HMAC integrity, Kubernetes RBAC escalation detection (kaaval, EGC)",
"📡 MCP Ecosystem" : "MCP server tooling, registry tests, Continue.dev integration, AI agent infra",
"🧪 Testing & Reliability" : "Unit tests, silent-skip detection, streaming parsers, CI hardening",
"🌍 Open Source" : "84 merged PRs across 18+ repos — features, security, fixes, refactors, docs",
}
Open to collaborating on · AI/LLM tooling · MCP integrations · Security engineering · Kubernetes · Python open source · Android