341 malicious skills in ClawHub
_Extracted from: how-not-to-run-your-agents/index.md_
adelzaalouk
Personal blog of Adel Zaalouk
Latest posts
Across 40,000 game runs
_Extracted from: how-not-to-run-your-agents/index.md_
Anthropic's eval breach
_Extracted from: how-not-to-run-your-agents/index.md_
banking AI agent was compromised through a 0.01 EUR transfer
_Extracted from: how-not-to-run-your-agents/index.md_
hacked HuggingFace's production infrastructure
_Extracted from: how-not-to-run-your-agents/index.md_
LiteLLM versions 1.82.7 and 1.82.8 were compromised
_Extracted from: how-not-to-run-your-agents/index.md_
Pillar Security demonstrated
_Extracted from: how-not-to-run-your-agents/index.md_
How NOT to Run Your Agents (and What to Do Instead)
An OpenAI agent escaped its sandbox and hacked HuggingFace for nine days. Anthropic's Claude published a malicious PyPI package during evals. Cursor and Codex got compromised through workspace configs. These are the first six months of agents in production. Here are 21 anti-patterns to avoid.
Fifth Third Bank
_Extracted from: how-agents-run-in-production/index.md_
PayPal autonomous SRE
_Extracted from: how-agents-run-in-production/index.md_
How Agents Run in Production
The industry does not have a shared vocabulary for agent execution. Six execution modes and four scale archetypes give you a framework for deciding what you actually need to schedule, isolate, and sandbox these workloads.
Agentic Code Review
_Extracted from: your-code-review-process-is-already-broken/index.md_
AI Engineering Report 2026: The Acceleration Whiplash
_Extracted from: your-code-review-process-is-already-broken/index.md_
AI Tool Impact on Developer Productive Output
_Extracted from: your-code-review-process-is-already-broken/index.md_
BitsAI-CR: Two-Stage Code Review at ByteDance
_Extracted from: your-code-review-process-is-already-broken/index.md_
CodeRabbit's analysis of 470 open-source PRs
_Extracted from: your-code-review-process-is-already-broken/index.md_
HighStakes on GitHub
_Extracted from: your-code-review-process-is-already-broken/index.md_
Semantically-Seeded Impact Analysis
_Extracted from: your-code-review-process-is-already-broken/index.md_
The Verification Bottleneck: Why AI's Real Cost Is Human Attention
_Extracted from: your-code-review-process-is-already-broken/index.md_
HighStakes: where humans review, where AI handles the rest
Not all code changes carry the same risk. HighStakes scores every file by blast radius so your senior engineers review the code that matters and AI handles the rest.