AI Coding Agent Security: 10 Rules From Real Incidents
From PHP2Shell to LiteLLM and GitSpawn: 10 rules for your laptop, your CI pipeline, and your servers, each with a test to run today and the fix.
From PHP2Shell to LiteLLM and GitSpawn: 10 rules for your laptop, your CI pipeline, and your servers, each with a test to run today and the fix.
Latent Sparse Attention (LSA) from Prefrontal AI and ETH Zurich clusters keys to speed long-context LLM inference, cutting 10M-token time to first token by over 9 times
Two Hacker News posts say a team of Qwen3.8-27B models matches or beats Claude Fable 5. Here are the four reasons a second agent helps, and how to test yours.
Google's EnvHarness adapts existing agent training environments based on failure patterns, boosting ALFWorld held-out task success by nine percentage points
One shared error pulled three of OpenAI's math papers. Write the sentence your check proved under the claim you meant.
Timestral's largest model has 209 million parameters and leads three pretrained, zero-shot forecasting benchmarks. The smallest forecasts 96 steps from 2,048 observations in 17 milliseconds on eight CPU threads.
Ollama picks the window, flash attention, and an MTP draft before you type. Here is the one number that judges each, on a 24GB card and a Mac.
Jev and a local model matched my reference on 12 of 12 Claude Code fixes to a synthetic API. A filename rule matched 3 and still caught the one real bug!
Mem0 leads on GPT-4o-mini. Pi with Claude Opus 5.5 scores 99% for $6.95, against 95% at $17.84 for full context
A Claude Code fix passed 9 required tests and still returned the wrong title. One closed question after the tests picked the suite that caught it
Rough concept first, then the words. 10x to 31x faster than strong diffusion models, on 4x to 5x fewer training tokens
Your agent harness is an architectural decision that directly dictates task success and API costs. Here is how to match it to your workload.
How prompt injection exploits trust boundaries through saved memory, and what runtime checks can prevent unsafe actions.
Nvidia's Qwen3 14B to 32B mapper is 25.1x vs re-prefill at 32,768 tokens. Cache-to-Cache reports about 2.5x against text
Classify depth, time, or flow. Nanbeige is the downloadable LLM. RLT lost at 140M. Looped Flows is a 5-7M puzzle net
Self-Harness reports relative gains of up to 132%. EnvHarness adds up to 9.0 points and 9.8% fewer steps