TESSFeed

Technical AI signals · Daily at 8 PM ET

Friday, July 17, 2026

Security and Control Move Forward

Today was about operationalizing agents: OpenAI pushed Codex deeper into vulnerability repair, uv added OSV-backed malware blocking, GitHub opened up Copilot’s engine as an SDK, and NVIDIA expanded both its fine-tuning and embedding stack.

🛰️  Top Signals

1

OpenAI pushes Codex into vuln repair

OpenAI is positioning Codex Security as a real code-security workflow, with GPT-5.6 Sol aimed at finding, validating, and fixing vulnerabilities instead of just suggesting patches.

@OpenAI

3

GitHub exposes Copilot as an agent SDK

GitHub is productizing the Copilot runtime across six languages, giving teams a standard engine for planning, tool use, and file edits inside their own apps and workflows.

GitHub

X / Twitter

Blogs

The cost of saying yes has changed

GitHub frames a decision model for AI-era code changes around code getting cheaper to write while ownership cost does not.

GitHub Blog

A list of existing alignment approaches

The post organizes alignment techniques across dimensions including internals versus outputs, SFT versus RL, and online training versus toy-domain training.

LessWrong

Quoting Kimi K3

Simon Willison quotes Kimi K3 replying, "Is there something I can actually help you with today?" after refusing to leak its system prompt.

Simon Willison

HN discussion

Hacker News

Research

YouTube

Hugging Face

thinkingmachines/Inkling

Inkling is an open-weights multimodal model that takes text, image, and audio inputs and generates text for agentic, tool-use, coding, and RAG applications.

model

zai-org/GLM-5.2

GLM-5.2 is presented as a flagship long-horizon model with a 1M-token context plus coding with multiple thinking effort levels.

model

tencent/Hy3

Hy3 is a 295B-parameter MoE model with 21B active parameters and 3.8B MTP layer parameters, with Tencent saying it was post-trained using feedback from 50+ products.

model

OpenMOSS-Team/MOSS-Transcribe-Diarize

This 0.9B model handles transcription, diarization, timestamps, and acoustic events across 50+ languages in a single pass for recordings up to 90 minutes.

model

ATH-MaaS/OvisOCR2

OvisOCR2 is a 0.8B page-level document parser that outputs Markdown in reading order for text, formulas, tables, and visual regions, with a reported 96.58 OmniDocBench score.

model

Cactus-Compute/needle

Needle is a 26M-parameter pure-attention encoder-decoder distilled from Gemini 3.1, with claimed production speeds of 6000 tokens/sec prefill and 1200 decode.

model

InternScience/Agents-A1

Agents-A1 claims trillion-parameter-class performance from a 35B agent model, and the repository notes a new 4B release plus quantized variants.

model

openbmb/UltraX-Preview

UltraX publishes five English pretraining corpora of about 20B tokens each, using a refinement model to predict structured insert, delete, and modify edits that are then executed deterministically.

dataset

Glint-Research/Fable-5-traces

The dataset contains 4,665 Pi trace sessions converted from 60 source sessions, with 3,799 tool actions, 866 assistant text actions, and median reasoning length of 2,365 characters.

dataset

GitHub

PostHog/posthog

PostHog pitches an open-source product stack with a self-driving mode that turns product signals like errors, rage clicks, and failed queries into reports and pull requests.

Python · ★ 36,178

openinterpreter/openinterpreter

Open Interpreter says it reimplemented the provider-recommended Kimi Code harness in Rust and can switch harnesses with `/harness` for low-cost models.

Rust · ★ 66,346

tirth8205/code-review-graph

The tool builds a Tree-sitter structural map, tracks changes incrementally, and feeds graph-aware MCP context so coding agents read only the changed parts they need.

Python · ★ 19,732

anthropics/cwc-workshops

The workshop materials include model sweeps over quality-per-dollar and quality-per-second, plus a 400-line prompt decomposed into skills, code execution, and callable agents.

TypeScript · ★ 1,579

Nutlope/hallmark

Hallmark is a design skill for Claude Code, Cursor, and Codex with 20 themes, 4 verbs, 57 slop-test gates, and a pre-emit self-critique step.

CSS · ★ 11,990

RyanCodrai/turbovec

turbovec says a 10 million document corpus drops from 31 GB as float32 to 4 GB, with online ingest and SIMD search that beats FAISS by 10–19% on ARM in cited configs.

Python · ★ 13,292

PrismML-Eng/Bonsai-demo

The demo runs Bonsai 1-bit and ternary models locally across Metal, CUDA, Vulkan, ROCm, or CPU, and the new 27B line adds vision, OpenAI-style tool calls, MCP servers, and reasoning modes.

Shell · ★ 1,705

HKUDS/DeepTutor

Recent releases add selective removal of one failed document from a knowledge base and multimodal image extraction during LlamaIndex ingestion.

Python · ★ 27,344

Get this in your inbox

The same feed, delivered daily at 8 PM ET. No spam, unsubscribe anytime.