pokitdog.com
Show HN: Groundctl – "Terraform plan" for your local developer machine
Hi HN,I built groundctl because every team I've been on has the same problem: "it works on my machine." Someone upgrades Node, someone has the wrong Terraform version, a new hire spends two days debugging a setup that differs from everyone else's.Existing tools don't solve this. chezmoi manages dotfiles, not tool versions. mise/asdf manage language runtimes per-project, not team-wide standards. Ansible is overkill for local machines.groundctl is a single Go binary t
Ask HN: How are you managing secrets with AI agents?
Secrets management with Agents feels absent today. The agent needs API keys to call external services, but the usual patterns feel broken in this context. You see this clearly when writing Agent Skills.Environment variables: The agent has shell access. It can run `env` or `echo $API_KEY` and access the secret, either through prompt injection or just by exploring or debugging..env files: Same problem. The agent can `cat .env`. The file is right there on the filesystem waiting for curious `print()
Show HN: A user daemon to provide an age-bracketing API
(Before you react: I think these laws are pointless and dumb too, but they're still laws so here we are)Hi all, I saw a lot of talk about the new California (and pending Colorado) laws about requiring operating systems to provide an API to return a user's age bracket to applications. While I think most people agree that this is asinine, pointless, full of holes, open to abuse, etc. etc., it's also a legal requirement anyway.I've been playing with Claude Code lately and so I t
Feedback on evolutionary multi-agent architecture for nonstationary environments
Hi HN,
I’m experimenting with a systems architecture rather than a finished product, and I’d really appreciate technical feedback from people working on ML systems, agents, or complex adaptive systems.
The core idea is to treat strategies/agents as evolving “species” instead of static models. Each species is composed of modular “organs” (signal processors, risk controllers, execution logic, etc.), and mutations are constrained to compatible organ classes so components can recombine safely.
Show HN: SurvivalIndex – which developer tools do AI agents choose?
We've been running coding agents against standardized repos
with natural-language prompts — no tool names, no hints —
and measuring what they actually choose.Early finding: Claude Code picks Custom/DIY in 12 of 20
categories. Not because it can't use the tools (BFCL scores
suggest it can) but because it doesn't reach for them.
That's a different failure mode than capability benchmarks
measure.We score each tool on: agent visibility, pick rate vs
Custom/DIY, c
Show HN: Self-Evolving Skill – empirical results from a 5-round experiment
Last week I shared the design pattern. This week I ran a real experiment to validate it.Five rounds on a MySQL database (29 tables, 590MB). Key results:- Five-Gate rejection rate: 63.6% — most interactions produce no knowledge change
- Incremental convergence: +75 → +46 → +12 → +21 → +1
- Gate 2 self-correction: caught and fixed 2 erroneous rules the Skill had written earlier
- Round 5: zero exploration steps, direct template reuse
- Accuracy: 100% (no incorrect knowledge survived)Unexpected fin
Show HN: Harden – 5 AI models audit your code, then debate each other's findings
I built harden because I kept copy-pasting code between ChatGPT, Claude, and Gemini trying to cross-check their reviews. Each one found things the others missed, but synthesizing their outputs manually was painful.harden runs 5 frontier models (Claude, GPT-4o, Gemini, Mistral, DeepSeek) in parallel on the same input. They analyze independently, then cross-examine each other's findings. A coordinator synthesizes the debate into consensus findings and produces a fixed version.The key insight:
OpenAI might end up on the right side of history
note: I am in MENA, am not with the military in any way.when i first read the statement by Dario, i was shocked by the fact the military was so dismissive about Ai safety (not to mention privacy). Seeing anthropic resist the military, I felt so proud of being a claude user to the point I deleted gpt right away. it's nice to see your fav products sync with your values.but today, after thinking more about it, i realized something. for a government to allow one Ai company to dictate terms, it
Show HN: An offline-first nuclear survival iOS app–EPA RadNet,140 stations
This app prepares you to survive during nuclear wars. Motivation: save people.<p>Core Feature(coming soon): offline chat with nearby iOS devices using Multipeer Connectivity Framework.<p>1. Disaster Alerts: API from authorities like NOAA + FEMA IPAWS
2. Live Radiation Map: EPA RadNet, 140 official U.S monitoring stations network, all 50 states
3. Offline-first: Built to function when there's no internet
US service member dies of injuries from Iranian strike on Saudi Arabia
A U.S. service member passed away Saturday night, days after being wounded in an Iranian attack on American forces in Saudi Arabia, U.S. Central Command said.CENTCOM announced the death of the service member on Sunday afternoon, saying that the American was “seriously wounded at the scene of an attack on U.S. troops in the Kingdom of Saudi Arabia on March 1.” They are the seventh American service member to die during Operation Fury, the U.S. military’s name for combat operations against Iran.Acc
US strikes continue as Iran war enters its second week
Nine days into the war with Iran, U.S. forces alongside the Israeli military continue to bomb targets around Iran and throughout the Middle East as the conflict shows no signs of ending.U.S. forces have hit more than 3,000 Iranian targets as part of Operation Epic Fury, U.S. Central Command said, with fighter jets, bombers and missiles fired from destroyers. This weekend, President Donald Trump posted to social media saying that the U.S. military is expanding its range of targets and that Iran &
Claude Code Open Source?
Today I learned that they ship the entire CLI with the agent SDK. It's all there...Do we think this was a mistake?
----node_modules/@anthropic-ai/claude-agent-sdk/cli.js. It's the bundled, minified CLI binary for Claude Code itself (the tool you're talking to right now).
Key details:
- Version: 2.1.71, built on 2026-03-06
- ~13,800 lines of heavily minified/bundled JavaScript
- Shebang: #!/usr/bin/env node — it's meant to be run a
BaZi – Deterministic life-charting from the Chinese calendar
I've been fascinated by BaZi (八字), a Chinese calendrical system that maps
birth date/time to Five Element interactions. Unlike Western astrology, it's
essentially a deterministic lookup + combinatorial analysis — same inputs
always produce the same chart.
The core algorithm converts Gregorian dates to the Chinese Sexagenary cycle
(天干地支), then derives Four Pillars (year/month/day/hour), each a pair of
Heavenly Stem + Earthly Branch. From there it compu
Show HN: OrangeWalrus, an aggregator for trivia nights (and other events) in SF
Two problems I encountered personally:1) Some buddies and I went to a trivia night late last year, only to arrive to find it cancelled (with signs still on the walls saying it happened every Tuesday, etc)2) Sourcing ideas for fun things to do in the city on a given night, in a given neighborhood. Some sites help a ton (e.g. funcheapsf), but often don't have everything I'd want to see, so we decided to build that out a bit.Anyway, I built this originally to solve #1, then a buddy and I
Distinguish skipped CI from failed CI on PRs page
GitAuto's PR dashboard no longer shows "Failed" when CI was just skipped or hadn't run. Now it shows the real status from GitHub - Skipped, Timed Out, Cancelled, etc. Less noise when triaging PRs.
Show HN: Mtb – An MCP sanity checker for vibe coding
I originally conceived of Make the Bed (named after the Calvin & Hobbes strip where Calvin spends all day building a bed-making robot that never works instead of making his bed) as a tongue-in-cheek response to some of the vibe coded projects I've seen recently. To my surprise it works as intended and prompts users to consider many factors (existing solutions, maintenance costs, etc.) before starting a new project or feature. It also shows complexity metrics via scc.Check out the demo p
Amazon’s $4 multitool pocket knife is a clever and useful everyday carry tool
Amazon’s $4 multitool pocket knife is a clever everyday carry tool that fits on a keychain and handles small cutting tasks and more with ease.
Rough Edge EDC: Olitans King Penguin nearly perfect for everyday carry
EL PASO, Texas (KTSM) – A few months ago, I had never heard of Olitans. Now, I am a huge fan of this knife brand. My first Olitans was the Mini Knife, also identified by the model number T019. This knife is small, but super well constructed. Perfect for a fifth pocket or a secondary […]
15 Amazon gadgets to make your EDC smarter and more tactical
Tactical gadgets on Amazon blend cutting-edge tech with battlefield-tested designs, delivering reliable performance for urban emergencies and outdoor adventures.
My 5 essential EDC items for daily preparedness
A personal breakdown of five essential everyday carry items that support preparedness without overcomplicating daily life.