cd /news/artificial-intelligence/agent-harness-hackathon-2026-what-i-… · home topics artificial-intelligence article
[ARTICLE · art-115637] src=dev.to ↗ pub= topic=artificial-intelligence verified=true sentiment=↑ positive

Agent Harness Hackathon 2026 — What I built, how I used TrueForge, and what I learned

A developer built AutoVault, an autonomous AI security operations agent using TrueFoundry's open-source TrueForge harness, which detects ransomware in 0.3 seconds and spawns parallel subagents for investigation while requiring human approval for irreversible actions. The project, entered in the Agent Harness Hackathon 2026, leverages all 12 TrueForge features including MCP tools, subagents, sandboxing, and persistent sessions, and incorporates 13 innovations such as predictive defense and self-healing file systems.

read3 min views7 publishedAug 30, 2026

THE PROBLEM

Ransomware attacks cost organizations over $20 billion every year. Security teams are drowning in 11,000+ alerts per day, most of which are false positives. By the time a human analyst investigates, the encryption has already spread to hundreds of files.

The average response time? Hours. The average recovery cost? $2.73 million.

I wanted to build something that could change that — an AI agent that doesn't just detect threats, but investigates and responds to them autonomously, while still keeping a human in the loop for irreversible actions.

That's how AutoVault was born.

WHAT I BUILT

AutoVault is an autonomous AI security operations agent that runs on TrueForge, the open-source agent harness by TrueFoundry. It detects ransomware in 0.3 seconds, spawns 5 parallel subagents to investigate, and responds autonomously — but always stops to ask a human before doing anything irreversible.

HOW I USED TRUEFORGE

TrueForge isn't just a dependency — it's the brain of the entire system. I used every single TrueForge feature, all 12 of them.

MCP TOOLS

I built 4 Python MCP servers using the MCP SDK v2:

Each tool has full input schemas, proper error handling, and returns structured JSON.

SUBAGENTS

TrueForge's subagent system was a game-changer. I spawned 5 specialized agents that run in parallel:

The speed improvement was dramatic — what took hours now takes seconds.

SANDBOX (DAYTONA) I wrote 20 Python scripts that run safely in TrueForge's Daytona sandbox:

HUMAN APPROVALS

TrueForge's approval system lets me define 10 rules with risk scoring:

The 7-step workflow ensures nothing irreversible happens without a human saying yes.

SKILLS

I created 4 instruction packs that TrueForge loads dynamically:

CONTEXT ENGINEERING

TrueForge's context engineering kept the agent efficient:

PERSISTENT SESSIONS

The agent saves its state to SQLite every 30 seconds. If the connection drops mid-investigation, it resumes exactly where it left off.

THE 13 INNOVATIONS

Human-Like Memory Architecture — Episodic, semantic, procedural, and working memory. The agent remembers past attacks.

Predictive Ransomware Defense — Predicts attacks 30 minutes before they happen.

Recursive Self-Improvement — The agent improves its own code and algorithms.

Collective Intelligence Swarm — Hive mind of agents making decisions together.

Neuromorphic Processing — Brain-inspired neural networks with sub-millisecond latency.

Self-Healing File System — Automatically restores encrypted files from blockchain backups.

AI Red Team vs Blue Team — Two AI teams compete and improve through adversarial training.

Knowledge Graph — Graph database mapping relationships between threats, actors, tools, techniques.

Federated Learning — Multiple organizations train models without sharing sensitive data.

Digital Twin Simulation — Virtual replica for risk-free security testing.

Adversarial Robustness — Defends against attacks designed to trick the AI.

Natural Language Threat Intelligence — Query threats using plain English.

Self-Learning Security Evolution — Detection models evolve using genetic algorithms.

THE QODO CODE REVIEW EXPERIENCE

Every change goes through Qodo review via GitHub Actions. On PR #1, Qodo found:

All three were real issues. I fixed all HIGH findings and documented everything in the README.

WHAT I LEARNED

TrueForge Is More Than a Wrapper — The MCP tool system, subagent orchestration, approval workflows, and context engineering are all deep, production-grade features.

Parallel Subagents Change Everything — Running 5 investigations simultaneously isn't just faster — it's qualitatively different.

Human Approval Is Non-Negotiable — For a security tool, having human approval is critical. TrueForge made this trivial to implement.

Qodo Actually Finds Bugs — It understands context and finds issues that surface-level analysis misses.

The Sandbox Is Your Safety Net — When running scripts that analyze potentially malicious files, you need a sandbox.

Documentation Matters — Writing the README forced me to think about what I actually built and why.

TRY IT YOURSELF github.com/ItsParthPinjarkar/autovault

git clone https://github.com/ItsParthPinjarkar/autovault.git cd autovault

pip install -r agent/requirements.txt

pip install -r mcp-server/requirements.txt

bash setup-trueforge.sh

python demo_advanced.py

ACKNOWLEDGMENTS

Built for the Agent Harness Hackathon (August 24-30, 2026). This project was built with AI assistance. All code has been reviewed, tested, and understood by the developer.

── more in #artificial-intelligence 4 stories · sorted by recency
── more on @autovault 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/agent-harness-hackat…] indexed:0 read:3min 2026-08-30 ·