Piekwerk - DEV Community

Piekwerk

Nov 9

Alibaba measured it: the best coding agent meets 70% of requirements and finishes 26% of tasks

#ai #devtools #programming #testing

1

5 min read

Piekwerk

Nov 9

Microsoft's APM is npm for agent context: one apm.yml, 9 harnesses, byte-exact lockfiles

#ai #microsoft #devtools #programming

1

6 min read

Piekwerk

Nov 9

5,000 agents, 7,200 pushes a second: inside GitHub's Rubicon, and the Git guarantee it drops

#ai #git #github #devtools

5 min read

Piekwerk

Nov 8

OpenCode 2.2.6 pools your MCP connections: what pool:true saves, and what it can't

#ai #devtools #mcp #opensource

1

5 min read

Piekwerk

Nov 8

Google's Elephant-Goldfish plugin: 5 stages of fresh-context review before your agent writes code, and 3 catches

#ai #devtools #claude #programming

1

5 min read

Piekwerk

Nov 7

Anthropic's agent memory study: fixed save-rules cost Claude its gains, and 9 files beat 413

#ai #claude #programming #productivity

1

7 min read

Piekwerk

Nov 7

Claude Code's don't-ask-again file held 1,379 secrets in 1,122 public repos: what settings.local.json really saves

#ai #security #claude #devtools

1

5 min read

Piekwerk

Nov 7

MCP Inspector's OAuth flow could run PowerShell on Windows: inside CVE-2026-119025, and why npm update misses the fix

#ai #mcp #security #devtools

2

6 min read

Piekwerk

Nov 6

Claude Code background subagents now use 5% less of your usage limit: how the 99% status-line fix works

#ai #claude #devtools #productivity

1

4 min read

Piekwerk

Nov 6

AGENTS.md generators vs validated kits: what a 1,280-rule generator can't check

#ai #devtools #programming #productivity

1

6 min read

Piekwerk

Nov 5

Positron 2026.10 ships an IDE-wide MCP server: 5 guardrails before Claude Code gets your consoles

#ai #mcp #claude #devtools

2

5 min read

Piekwerk

Nov 4

Google migrated 174,000 lines of Fortran to Rust in 59 minutes: the 11-minute baseline is the part to copy

#ai #rust #devtools #programming

1

6 min read

Piekwerk

Nov 4

Google's Jules shuts down December 31: 7 days left to export your coding agent's memories and setup

#ai #devtools #programming #productivity

1

5 min read

Piekwerk

Nov 4

MCPhantom: agents chose deprecated MCP servers in 98% of tracked sessions, and no patch is coming

#ai #mcp #devtools #programming

2

5 min read

Piekwerk

Nov 3

Claude Code 2.1.317 tightens failure defaults: SDK hook callbacks can block, and an email leak now asks first

#ai #claude #devtools #programming

2

5 min read

Piekwerk

Nov 2

Vercel agent-desktop: computer use for coding agents through the accessibility tree

#ai #devtools #programming #opensource

1

5 min read

Piekwerk

Nov 1

Microsoft's TokenEconomizer cuts coding agent costs up to 30.9%, and it does it by editing your config files

#ai #devtools #programming #opensource

1

5 min read

Piekwerk

Nov 1

DeepSeek's dsh-superagent is a proactive coding agent that waits to be asked: router, tickets, and the observe default

#ai #devtools #programming #opensource

1

5 min read

Piekwerk

Oct 31

Programmatic tool calling goes default in Droid 1.0: one program, hundreds of tool calls, and a new allowed_callers decision

#ai #devtools #programming #productivity

1

6 min read

Piekwerk

Oct 31

Copilot Fusion is GA: 2 points off Opus 5.5 Max at half the cost, and your model picker is now an effort dial

#ai #github #devtools #programming

1

5 min read

Piekwerk

Oct 31

Claude Code 2.1.316: subagent effort becomes a setting, ultrathink gets a kill switch, and Opus fast mode has 30 days left

#ai #claude #devtools #programming

1

5 min read

Piekwerk

Oct 30

Malware posed as Claude Code in Cursor's marketplace: what your agent config can and cannot contain

#ai #security #devtools #claude

1

5 min read

Piekwerk

Oct 27

Cursor's Agent SDK is GA: your agent is a folder of Markdown and TypeScript

#ai #cursor #devtools #typescript

1

4 min read

Piekwerk

Oct 27

GitHub Copilot's assisted permissions went GA: an LLM now decides which commands run without asking you

#ai #github #devtools #security

1

4 min read

Piekwerk

Oct 26

MCP tasks landed in Claude Code: a tool call can now outlive your session, and 6 config decisions come with it

#ai #mcp #devtools #programming

2

1

6 min read

Piekwerk

Oct 26

MCP C# SDK 2.0 is stable: HTTP is stateless by default, and 10 changes can break your upgrade

#ai #mcp #csharp #dotnet

1

5 min read

Piekwerk

Oct 25

A web page can plant AGENTS.md in your Downloads folder: Codex obeyed it in over 90% of attempts

#ai #security #programming #devtools

1

2

6 min read

Piekwerk

Oct 25

MCP servers v2.1.0 fixes 8 vulnerabilities: filesystem writes could escape allowed folders

#ai #mcp #security #devtools

1

3

6 min read

Piekwerk

Oct 25

Microsoft's Agensh runs 1,024 coding agents with no manager: one agent on the newer model beat the whole fleet

#ai #devtools #programming #opensource

1

6 min read

Piekwerk

Oct 24

Claude Code 2.1.309 and 2.1.310 patched 15 guardrail bypasses in two days: deny rules, sandbox, and fail-open hooks

#ai #claude #devtools #security

1

7 min read

Piekwerk

Oct 24

20 of 21 agent configs just scored lower: what the Coding Agent Index v2.0 reset measures

#ai #devtools #programming #testing

1

5 min read

Piekwerk

Oct 23

Agent Skills moved to the Linux Foundation: 4 things SKILL.md authors should watch

#ai #devtools #programming #productivity

1

5 min read

Piekwerk

Oct 23

GitHub's Copilot default policy is enforcing itself through October 26: what to check before it chooses for you

#ai #devtools #programming #productivity

1

4 min read

Piekwerk

Oct 23

Claude Code's new safety_event finally shows which guardrail blocked your agent

#ai #claude #devtools #programming

5

4 min read

Piekwerk

Oct 22

GitHub's PDD Kit proves agent-written Rust correct: 5 steps and 3 limits

#ai #github #devtools #programming

2

3

5 min read

Piekwerk

Oct 21

GitHub's package firewall went into preview: npm gets checked before your agent installs it

#ai #security #devtools #programming

1

2

5 min read

Piekwerk

Oct 21

OpenAI's 12-hour prompt cache is live: split your agent context into two TTLs

#ai #openai #devtools #programming

2

5 min read

Piekwerk

Oct 20

model: haiku stopped meaning one thing. Pin the exact model id.

#ai #claude #devtools #programming

1

3 min read

Piekwerk

Oct 20

Codex 0.161 flipped the default model to GPT-6.1 Sol. If you never set model, you already switched.

#ai #openai #devtools #programming

1

4 min read

Piekwerk

Oct 20

20 things AI coding agents get wrong: the config rules that stop them

#ai #programming #devtools #claude

1

4 min read

Piekwerk

Oct 9

Claude Haiku 5.5 is GA in GitHub Copilot: 4 config surfaces decide what runs your subagents

#ai #programming #devtools #productivity

1

4 min read

Piekwerk

Oct 9

Claude Code 2.1.295 Adds onFailure block: Hooks Can Fail Closed Now, If You Ask

#ai #claude #devtools #security

1

5 min read

Piekwerk

Oct 8

Claude Code 2.1.294 Fixed Guards That Silently Failed: Hooks That Allowed, Rules That Never Loaded

#ai #claude #devtools #security

1

2

5 min read

Piekwerk

Oct 7

Claude Code 2.1.292 Asks Your MCP Servers for 2026-07-28: I Probed the Official Servers, All Four Answered 2025-11-25

#ai #mcp #devtools #programming

1

5 min read

Piekwerk

Oct 7

Superpowers at 296k stars: I read all 15 skills, and the whole methodology is 170 KB of markdown

#ai #claude #devtools #opensource

5 min read

Piekwerk

Oct 7

Your AI agent's config files have secrets in the clear: a 22-rule scanner that finds them

#ai #security #devtools #programming

1

5 min read

Piekwerk

Oct 6

Hugging Face measured it: the same model scores 33% or 62% depending on the harness

#ai #machinelearning #devtools #programming

1

2

5 min read

Piekwerk

Oct 6

JetBrains Air puts 5 agents in one IDE: 4 config checks before you install the EAP

#ai #devtools #programming #productivity

1

4 min read

Piekwerk

Oct 6

NVIDIA measured it: coding agents scored 19% without a spec file, 100% with one

#ai #devtools #programming #opensource

1

4 min read

Piekwerk

Oct 5

DeepSeek Harness added a Claude Code Mods compatibility layer: it's a subset test, not a port

#ai #devtools #programming #opensource

1

4 min read

Piekwerk

Oct 5

Slack Code went live today: 5 agents, one channel, and the rules files that keep them aligned

#ai #productivity #devtools #programming

1

5 min read

Piekwerk

Oct 4

OpenAI's MCP Extensions put plugins in ChatGPT's sidebar: 13 features and the portability bill

#ai #openai #mcp #devtools

1

5 min read

Piekwerk

Oct 4

OpenAI shipped Codex plugins: your agent config surface just moved out of the repo

#ai #openai #devtools #mcp

1

5 min read

Piekwerk

Oct 4

Claude Code 2.1.289 patched 4 deny-rule bypasses: your npm channel decides whether you have them

#ai #claude #security #devtools

1

4 min read

Piekwerk

Oct 3

Pi 1.0 reversed on MCP and shipped Codemode: where tool composition actually runs now

#ai #mcp #devtools #programming

1

5 min read

Piekwerk

Oct 3

Cursor's self-hosted runners feel on-prem: here's where the agent loop still runs

#ai #cursor #devtools #devops

1

5 min read

Piekwerk

Oct 3

A 7-line Claude Code mod overrode a deny rule: the plugin audit that matters now

#ai #claude #security #devtools

1

5 min read

Piekwerk

Oct 2

Copilot can click your desktop now: 4 guardrail details before you type /computer on

#ai #githubcopilot #devtools #security

2

4 min read

Piekwerk

Oct 2

Claude Code 2.1.287 ships mods: what in-process hooks change for your rules and guards

#ai #claude #devtools #security

1

5 min read

Piekwerk

Sep 28

NVIDIA moves the agent boundary into the kernel: what OpenShell GA means for your rules file

#ai #security #devtools #programming

1

5 min read

4 Week Community Wellness Streak

2 Week Community Wellness Streak

1 Week Community Wellness Streak

Writing Debut

Want to connect with Piekwerk?