michabbb

#devop, #php, #laravel, #mysql, #api expert, #ebay #developer, #openai, #ai, #llm,. #livewire, #alpinejs, searchable

I built https://burner-note.link🔒⏳🔥

Public Key

npub1z20c8zvvwqydxdthrlngrm8e08nhv7ketrz49lu9m6tzhghhwklql84yd9

NIP-05 Address

[email protected]

Profile Code

nprofile1qqsp98ur3xx8qzxnx4m3le5panuhnemk0tv4332jl7zaa93t5tmht0spz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0dslxf7sd

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1z20c8zvvwqydxdthrlngrm8e08nhv7ketrz49lu9m6tzhghhwklql84yd9

Show more details

Last Notes

2026-04-16 09:42:20 GMT

- reply

@nprofile…0c9w???

2026-04-13 13:40:20 GMT

Hey #php developers out there 👀 👋
What tools do you use to make your #ai write "good" code? 😏
#phpstan, #psalm, #mago, cs-fixer, #coderabbit...? 🤔
What's your most important tool? 🤔 #coding
https://social-cdn.vivaldi.net/system/media_attachments/files/116/397/677/824/932/637/original/48d601e82628a014.mp4

2026-04-11 20:43:47 GMT

#OpenDataLoader PDF — #opensource #PDF parser for #AI & #accessibility 🚀 #1 in benchmarks (0.90 overall), runs 100% locally, zero cloud, no GPU needed. #RAG #LLM #Python #NodeJS #Java
📄 Extracts #Markdown, #JSON (with bounding boxes) & #HTML from any PDF — correct reading order, heading hierarchy, list & image detection with XY-Cut++ algorithm
🧵👇#pdf
https://social-cdn.vivaldi.net/system/media_attachments/files/116/388/018/456/021/663/original/9a8962025c07ad55.png

2026-03-13 11:06:36 GMT

Windows 11 finally has a real sudo command (since 24H2) 🚀
Run admin commands directly in the terminal without reopening PowerShell as Administrator. Small change — huge productivity boost.
#Windows11 #DevTools #PowerShell #Terminal #Developer #Productivity 💻⚡
https://www.xda-developers.com/windows-shipped-sudo-command-changes-how-i-use-terminal/

2026-01-24 00:38:39 GMT

- reply

🎬 Rich actions with confirmation dialogs, onSuccess & onError callbacks
👁️ Conditional visibility based on data, auth, or complex logic
📦 Two packages: @.json-render/core (types, schemas) + @.json-render/react (renderer, hooks)
🔧 Schema definition with #Zod for type-safe component props
📤 Export as standalone #React code with no runtime dependencies
https://github.com/vercel-labs/json-render

2026-01-24 00:38:39 GMT

📋 #jsonrender by #Vercel: AI-generated UI with guardrails #AI #React #opensource #webdev
🎯 Define a component catalog as guardrails - #AI can only use components you define
🔄 User prompts get converted to constrained #JSON that matches your schema every time
⚡ Streaming & progressive rendering as the model responds
🔗 Data binding with JSON Pointer paths for two-way binding
🛡️ Built-in validation with customizable checks and validateOn events
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/947/281/825/339/697/original/e878a9cb98690e23.png

2026-01-24 00:28:07 GMT

- reply

🎯 Zero accuracy loss - preserves what matters: errors, anomalies, high-scoring items & query-relevant content using BM25/embedding similarity
✅ Full provider support: #OpenAI, #Anthropic, #Google, #Cohere, #Mistral & #LiteLLM with exact token counting
📈 Performance: Search results (1000 items) 45K→4.5K tokens (90%), Log analysis 22K→3.3K tokens (85%), API responses 15K→2.25K tokens (85%)

2026-01-24 00:28:06 GMT

- reply

🔄 CacheAligner: Stabilizes prefixes so provider caching actually works - up to 10x savings by fixing dynamic content like dates that break cache hits
📊 RollingWindow: Manages context within limits without breaking tool calls - drops oldest tool outputs first, never breaks system prompts or recent turns
🔧 Easy setup options: Proxy server works with any #OpenAI-compatible client without code changes, or use #Python SDK for fine-grained control

2026-01-24 00:28:06 GMT

🧠 #Headroom - The Context Optimization Layer for #LLM Applications #opensource #Python
💸 Cut your #AI costs by 50-90% without losing accuracy! Smart compression for massive contexts from tool outputs, search results & API responses
⚡ SmartCrusher: Compresses tool outputs statistically (70-90% savings) - keeps errors, anomalies & relevant items while dropping redundant data
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/947/240/556/910/726/original/8266053e410be990.png

2026-01-24 00:21:22 GMT

- reply

📊 Unified observability with #OpenTelemetry distributed tracing integration & structured error codes providing complete auditable trail of every tool call
🔧 Robust resiliency features: deadlines, timeouts & automatic flow control prevent cascading failures from unresponsive tools
🌐 Polyglot development: generates code for 11+ languages maintaining consistent strongly-typed contracts across implementations

2026-01-24 00:21:21 GMT

- reply

🔄 Full duplex bidirectional streaming enables client and server to send continuous data streams simultaneously over a single persistent connection - real-time agentic workflows without application-level synchronization
🛡️ Enterprise-grade security built-in: mutual TLS for Zero Trust architectures, native JWT/OAuth authentication hooks, method-level authorization for least privilege principle

2026-01-24 00:21:21 GMT

#GoogleCloud announces #gRPC as native transport for #MCP #AI
#MachineLearning #opensource
🔌 Eliminates need for transcoding gateways between #gRPC services and #MCP JSON-RPC
⚡ Binary encoding via #ProtocolBuffers shrinks messages up to 10x compared to JSON - dramatically lower latency, reduced bandwidth consumption & faster serialization
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/947/214/031/480/805/original/46f13fc21d8b8a2f.png

2026-01-23 20:36:57 GMT

#PhpStorm 2025.3.2 is now available 👀
#php
https://youtrack.jetbrains.com/articles/WI-A-231736298
https://social-cdn.vivaldi.net/system/media_attachments/files/115/946/331/484/888/437/original/6036563193f36617.png

2026-01-23 19:08:55 GMT

🩻 I Tried to Use #AI to Read My Friend’s Mom’s #MRI 🩻
spoiler: don't trust Ai 😏
#medgemma #health #google #vertexai
https://macropage.medium.com/i-tried-to-use-ai-to-read-my-friends-mom-s-mri-ed642700e95d?source=friends_link&sk=6b081ac6739d3813e334cb697460b72d
https://social-cdn.vivaldi.net/system/media_attachments/files/115/945/985/510/382/606/original/b04a7d18f226bcbe.png

2026-01-21 12:54:11 GMT

- reply

i love it.... it´s so convenient.... 😊
#linux #devops
https://social-cdn.vivaldi.net/system/media_attachments/files/115/933/186/474/261/875/original/f83b402b3eb097b9.png

2026-01-21 11:33:36 GMT

Learn #ai (now) or find another job (now) ‼️
https://www.youtube.com/live/OzkO1o8q3_4
A #developer who uses chatgpt to ask some questions and thinks he is using Ai, didn't understand anything..... 😔
https://social-cdn.vivaldi.net/system/media_attachments/files/115/932/869/638/513/750/original/68179cdc0e470aa7.mp4

2026-01-21 10:38:32 GMT

WATCH LIVE: #Nvidia President #JensenHuang speaks at #Davos
https://www.youtube.com/live/__IaQ-d7nFk
#ai
https://social-cdn.vivaldi.net/system/media_attachments/files/115/932/653/477/191/001/original/8e4cbc69caa7c407.jpeg

2026-01-21 00:19:52 GMT

📋 #SkillSeekers - #Python tool automatically transforms documentation, #GitHub repos & PDFs into production-ready AI skills #opensource #AI #LLM
🌐 Multi-source support: Documentation websites, #GitHub repositories & PDF files as data sources
🤖 4 LLM platforms supported: #Claude, #Gemini, #OpenAI & generic Markdown export
⚡ Three-Stream Analysis: Automatically splits repos into Code (C3.x), Docs &Insights streams
📦 24 preset configs for popular frameworks like #React, #Vue, #Django
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/930/220/975/425/806/original/0a3646075a976b68.png

2026-01-21 00:19:52 GMT

- reply

✨ AI enhancement adds explanations, examples & best practices automatically
🔧 MCP integration with 27+ tools & fully automated pipeline
✅ 1200+ tests, MIT licensed, v2.7.0
https://skillseekersweb.com/

2026-01-17 16:20:42 GMT

😶‍🌫️ 🚀 Imagine every host notifies you in #Slack and gives you the option
to upgrade the packages with one click....
I call this: a #devops heaven 🤗 ☁️ 🌞
#linux #security
https://social-cdn.vivaldi.net/system/media_attachments/files/115/911/350/019/218/621/original/f400180da633029e.png

2026-01-17 15:59:44 GMT

🤖 testing G(get) S(hit) D(one) 💩 💪
it’s a token burner 🔥
But in the end, I have to agree: IT GETS SHIT DONE 😂 🚀
https://github.com/glittercowboy/get-shit-done
#ai #coding #claude #claudecode
https://social-cdn.vivaldi.net/system/media_attachments/files/115/911/267/721/025/945/original/904682b2100e1920.png

2026-01-16 20:28:28 GMT

- reply

@nprofile…96cz because nobody is paying a developer these days for writing code (anymore) - if i hire someone, that person needs to know how to orchestrate ai agents. asking someone about their AI stack shows which century they live in 😉

2026-01-16 20:25:05 GMT

- reply

@nprofile…96cz okay, thanks for the info ;)

2026-01-16 20:23:04 GMT

- reply

@nprofile…96cz claude code ? codex ? opencode ? AGENTS ? ;)

2026-01-16 20:20:22 GMT

🔔 #OpenCode CLI Plugins 👉 Very useful😏
Here’s a little example:
Audio notifications when permissions are needed!
Play system sounds on #Linux, #macOS & #WSL2
using the permission.asked event 🎵
#ai #coding
https://gist.github.com/michabbb/99372884e1b774ee6e63d4d3995a5696
https://social-cdn.vivaldi.net/system/media_attachments/files/115/906/630/117/318/540/original/b8e1699f6b1d58bd.png

2026-01-16 03:25:05 GMT

- reply

@nprofile…96cz hi, what's your ai stack?

2026-01-14 16:48:42 GMT

- reply

@nprofile…5z55
Exactly right about the idle server scenario! But here's the part I find most elegant about this approach:
The target server computes absolutely nothing.
With classic rsync --checksum, both source AND target need to calculate checksums for every file. That means:
- Double the CPU load (both machines working)
- Double the I/O operations (both machines reading all files)
- The target might be a busy production server that can't afford the load
With the BLAKE3 + --files-from approach:
- Source: calculates hashes once, compares locally
- Target: just receives the delta files via rsync
- No checksumming on target at all
So even if your source server isn't idle, you're only stressing ONE machine instead of two. The target could be a busy production box, a low-power NAS, or even a remote server with expensive I/O - it doesn't matter because it just receives files.
That's the real win for me: asymmetric workload distribution. All the intelligence stays on the source side.
And yes, perfect for cold standby re-sync scenarios - but also great for continuous sync to production targets that shouldn't be burdened with checksum calculations!

2026-01-14 16:20:39 GMT

- reply

@nprofile…pz8e what is tiger ? do you have a link ?

2026-01-14 16:19:56 GMT

- reply

@nprofile…5z55
Good question! The numbers are from a production server with NVMe storage.
Breakdown:
- 19k files (15 GB): ~8-9 sec
- 13k files (53 GB): ~13-14 sec
- Total: 32k files (68 GB): ~21 sec
That's ~3.2 GB/s throughput - achievable with:
1. NVMe SSDs (3-7 GB/s sequential read)
2. Linux page cache on subsequent runs
3. Parallel reads via find -print0 | xargs -0 b3sum

2026-01-14 15:43:11 GMT

- reply

💡 Why hashes over timestamps?
- Timestamps can be manipulated
- Backup restores mess up mtime
- Hashes reliably detect real content changes
🚀 Result: 32,000 files completely scanned in under 30 seconds – syncing only the delta!

2026-01-14 15:43:10 GMT

📦 Efficient file syncing for massive datasets? Forget classic rsync!
⚡ #BLAKE3 hashes instead of timestamps – hash 70GB in just ~20 seconds thanks to multithreading across all CPU cores
#DevOps #Linux #SysAdmin #OpenSource #FileSync
📊 Algorithm comparison for 70GB:
- BLAKE3: ~20 seconds
- MD5: ~1.5 minutes
- SHA256: ~4.5 minutes
🔍 The trick: Create a hash baseline, compare with `comm -13`, then use `rsync --files-from` to transfer only actually changed files
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/894/215/488/924/512/original/d09c0f9c858aa345.png

2026-01-14 15:30:41 GMT

job done. context left: 0 😏
That was spot on 😊
#ai #coding
https://social-cdn.vivaldi.net/system/media_attachments/files/115/894/166/480/021/631/original/d46881d11ca71ffc.png

2026-01-13 02:54:22 GMT

yes, #openai #codex is slow, but hell.... that thing stays on track doesn't matter how long the session gets.... had a super long session (>3 hours) and that thing never lost focus... not a single fu**ing time.... 😮
#ai #coding
https://social-cdn.vivaldi.net/system/media_attachments/files/115/885/530/137/888/060/original/a6eeb538ef091075.png

2026-01-13 00:45:00 GMT

- reply

⚙️ Automatic model selection: haiku for trivial, sonnet for standard, opus for critical planning
🐳 Flexible isolation: #Docker containers, #git worktrees, or direct execution
💾 Built-in crash recovery via #SQLite - resume multi-hour tasks exactly where they stopped
🔀 Custom workflows: define any agent topology (expert panels, staged gates, hierarchical)
📦 Install via #npm: npm install -g @covibes/zeroshot
⚡ Requires #NodeJS 18+, Claude Code CLI, and GitHub CLI

2026-01-13 00:45:00 GMT

- reply

https://github.com/covibes/zeroshot

2026-01-13 00:44:59 GMT

🤖 #Zeroshot creates autonomous engineering teams for #ClaudeCode #AI #DevTools #opensource
🔧 Point at a #GitHub issue, walk away, and return to verified, production-ready PRs
🛡️ Multiple isolated agents check each other's work - validators didn't write the code, so they can't lie about tests
🔄 Smart task classification: TRIVIAL (1 agent) → CRITICAL (7 agents with security, adversarial testing)
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/885/021/553/100/447/original/cc2a3352669f912a.png

2026-01-05 15:47:21 GMT

- reply

🔄 Multi-turn conversation testing - verify AI behavior over 10+ exchanges, catch persona drift and premature agreement issues
🛡️ Negative tests with data providers - define hard boundaries for things AI should never do: reveal system prompts, acknowledge being AI, break character
📊 Statistical robustness - run tests N times with pass rate thresholds (e.g., 80% of 10 runs must pass) to catch flaky prompts

2026-01-05 15:47:21 GMT

- reply

🔧 Prompt sensitivity testing - verify small phrasing changes don't break behavior (formal vs casual openings)
💰 Cost tracking built-in - use cheap model for conversation, smarter model for judging. Budget alerts at $5/month, run tests nightly not on every commit
🎓 Real catches: character breaks, premature agreement, prompt injection leaks, persona drift in long conversations
https://joshhornby.com/eval-testing-llms-in-phpunit

2026-01-05 15:47:20 GMT

📋 Eval #Testing #LLM Outputs in #PHPUnit - A comprehensive guide for #PHP developers building #AI applications
🎯 Core problem: Traditional unit tests assert exact outputs, but #LLMs produce probabilistic responses - same prompt yields different wording each time
⚖️ Solution: LLM-as-judge pattern - test behavior not strings. Use a smarter model to evaluate whether responses meet criteria written in plain English
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/843/271/265/941/208/original/cb300db96fa3b8b1.png

2025-12-31 00:23:29 GMT

Seriously, this is what every single man's toilet looks on this planet, and I don't get it - most people are stupid, okay, but this is another level... Crazy... How do they piss? From the other side of the room? Men, what's wrong with you? 🤦🙄
#wtf #men
https://social-cdn.vivaldi.net/system/media_attachments/files/115/811/326/839/070/189/original/1dbeea0be35c6df9.jpeg

2025-12-28 10:40:47 GMT

- reply

🚀 Real results: Geoffrey Huntley ran a 3-month loop building a complete programming language. YC hackathon teams shipped 6+ repos overnight for $297 in API costs
✅ Best for: Large refactors, batch operations, test coverage, documentation generation - tasks with clear completion criteria
⚠️ Not for: Ambiguous requirements, architectural decisions, security-sensitive code, or exploration work

2025-12-28 10:40:47 GMT

🔄 #RalphWiggum: Autonomous Development Loops for #ClaudeCode #AI #DevTools #Automation
🎯 Official #ClaudeCode plugin that turns AI into a persistent loop - give it a prompt, it works until done or you stop it
⚙️ Uses Stop hooks to intercept Claude's exit, re-feed the original prompt with each iteration seeing modified files and #git history from previous runs
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/796/767/344/915/938/original/063848927958a816.jpeg

2025-12-28 10:40:47 GMT

- reply

🔧 Install via: /plugin marketplace add anthropics/claude-code then /plugin install ralph-wiggum@claude-code-plugins
💡 Philosophy: "Better to fail predictably than succeed unpredictably" - prompt engineering becomes the critical skill
https://paddo.dev/blog/ralph-wiggum-autonomous-loops

2025-12-25 22:05:59 GMT

- reply

🚀 Enable via new Plugins section at https://openrouter.ai/settings/plugins - opt-in toggle activates automatic healing layer for all structured output requests
💡 Key insight: A 2% defect rate dropping to 1% means cutting defects, bugs, and support tickets in half - reliability at the margins is where production systems succeed or fail
https://openrouter.ai/announcements/response-healing-reduce-json-defects-by-80percent

2025-12-25 22:05:58 GMT

- reply

• Type Coercion: 0.041ms (32,600 ops/second)
• Basic Parsing: 0.133ms (16,900 ops/second)
• Large Payloads (10KB): 2.3ms (437 ops/second)
💰 Cost: Free plugin with negligible latency impact (<1ms for typical responses)
🎯 Scope & limitations:
• Fixes JSON syntax errors, not schema adherence
• Works for non-streaming requests only
• XML healing available on request
• Tool calling has few structural issues, schema adherence evaluation coming soon

2025-12-25 22:05:58 GMT

- reply

• #Mistral Small 3.2: 99.3% defect reduction (98.8% → 99.99%)
• Ministral 3B: 100% defect reduction → perfect 100% validity
🛠️ Common issues fixed automatically: trailing commas after last element, unescaped control characters in strings, missing closing brackets, various syntax errors that break JSON parsers, prefixed text like "Here's the data you requested:"
⚡ Performance benchmarks:
• Schema-less Repair: 0.018ms (54,700 ops/second)
• Unified API: 0.019ms (51,500 ops/second)

2025-12-25 22:05:57 GMT

🔧 #OpenRouter launches Response Healing: automatically fixes malformed #JSON responses from #LLMs before they reach your app #AI #API #devops
📊 Results from 1.6M+ requests analyzed at inference time:
• #Gemini 2.0 Flash: 80% defect reduction (99.61% → 99.92%)
• #Qwen3 235B: 99.8% defect reduction (88% → 99.98%)
• #Deepseek Chat v3.1: 85% defect reduction (82.5% → 97.4%)
• #Devstral 2512: 99.6% defect reduction (96.6% → 99.99%)
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/782/474/521/625/925/original/0b7b9253cdc4056b.png

2025-12-25 17:35:28 GMT

🚀 #PHP 8.6 introduces Partial Application
Functions can be partially applied, returning a Closure where remaining parameters are supplied later. This reduces boilerplate, simplifies callbacks and supports cleaner functional programming patterns 💡
More details: https://thephp.foundation/blog/2025/12/08/partial-application/
#PHP #PHP86 #WebDev
https://social-cdn.vivaldi.net/system/media_attachments/files/115/781/410/910/861/019/original/a7022fc1e7acfc02.png

2025-12-25 17:26:23 GMT

#php pipe operator: do you remember the hype (some wannabe influencer created) ❓
https://dev.to/tegos/php-85-pipe-operator-is-it-worth-using-4gig
https://social-cdn.vivaldi.net/system/media_attachments/files/115/781/375/170/094/503/original/938737e299b77f42.png

2025-12-25 17:15:42 GMT

- reply

https://dev.to/laravel_mastery_ffd9d10ec/php-killed-dynamic-properties-heres-why-and-what-to-do-3knh

2025-12-25 17:15:42 GMT

- reply

⚡ Escape hatch exists: #[AllowDynamicProperties] attribute opts into old behavior. Use ONLY for legacy vendor code or gradual migration with a removal plan. Never use for new code.
📊 Timeline: PHP 8.1 = allowed | PHP 8.2 = deprecated warning | PHP 9.0 = fatal error. Start fixing now!
🎯 Migration strategy: Run PHPStan/Psalm analysis, fix domain models first (critical), then controllers/services, then DTOs. Add static analysis to CI pipeline to prevent regression.

2025-12-25 17:15:41 GMT

- reply

🔓 Problem 2: Zero type safety. Your age property can be int, string, null, array - PHP accepts anything without complaints. Type errors only surface at runtime, often in production.
🛠️ Problem 3: IDEs and tooling are blind. No autocomplete suggestions, no refactoring support, static analysis tools like PHPStan and Psalm cannot detect issues. You lose all modern developer experience benefits.

2025-12-25 17:15:41 GMT

- reply

🏗️ Problem 4: Architecture decay. Classes become dumping grounds for random data: $user->cached_data, $user->temp_flag, $user->random_stuff. Code becomes unmaintainable.
✅ Modern PHP solution: Declare all properties explicitly with types. Use public string $name; public int $age; public bool $isActive; in your class definitions.

2025-12-25 17:15:41 GMT

- reply

🚀 PHP 8+ feature: Constructor property promotion simplifies everything. Write public function __construct(public readonly string $customerId, public readonly array $items) {} - clean, typed, immutable.
💎 #Laravel specific: Use $fillable arrays and $casts for Eloquent models. Use Attribute accessors for computed properties. Never add undefined properties directly to models.

2025-12-25 17:15:40 GMT

⚠️ #PHP 8.2 deprecated dynamic properties - #PHP9 removes them completely! #webdev
🔴 Dynamic properties = adding properties to objects without declaring them in the class. Example: $user->age = 30; on an empty User class. This worked silently in older PHP versions.
🐛 Problem 1: Silent typos become hidden bugs. Write $user->custmer_email instead of customer_email - no error thrown, just returns null later when you try to read it. Hours of debugging for a simple typo.
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/781/333/181/265/631/original/ed3c88746320d7f0.png

2025-12-25 17:06:44 GMT

The fact that the word #Docker isn't used anywhere on this page is fucking insane!
wtf are people using out there?
https://xdebug.org/funding/001-native-path-mapping
Anyway, in case you didn't notice: #php #Xdebug (finally) supports path mapping, something every tool should provide!
https://social-cdn.vivaldi.net/system/media_attachments/files/115/781/297/929/574/115/original/9a7f26040eb0eb3e.png

2025-12-20 19:18:25 GMT

🔐 #SOPS: Secrets OPerationS - simple & flexible tool for managing secrets #devops #security #opensource #infosec
🔑 Encrypts #YAML, #JSON, ENV, INI & binary files with multiple encryption backends: #AWS KMS, #GCP KMS, Azure Key Vault, #age & #PGP
🛡️ Only encrypts values while leaving keys in cleartext for easy diffing & version control in #git repositories
🔄 Key groups with Shamir's Secret Sharing - split data keys requiring multiple identities to decrypt
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/753/504/136/859/466/original/d44c4549c7204b15.png

2025-12-20 19:18:25 GMT

- reply

⚡ Multiple modes: edit files in-place, encrypt/decrypt stdin/stdout, exec-env for secrets in environment variables
📋 Configuration via .sops.yaml with path-based rules for automatic key selection per environment (dev/prod)
🔗 Integrates with #HashiCorp Vault transit backend for enterprise key management
📊 Key rotation, auditing to PostgreSQL, and publishing to S3/GCS/Vault supported
⭐ 20k stars on #GitHub | MPL-2.0 license
🔗 https://github.com/getsops/sops

2025-12-20 19:08:36 GMT

(german) lustig... #chatgpt "selbst" nutzt es selten.... find ich ja cool das chatgpt in seiner freizeit #php nutzt 😏
#ai #fun
https://social-cdn.vivaldi.net/system/media_attachments/files/115/753/465/676/261/612/original/169ad4882706bfe8.png

2025-12-20 18:55:06 GMT

#DockerSecurityScanner - Containerized Security Scanning for Host Systems #security #docker #opensource #infosec #devops
🐳 All-in-one #Docker container that mounts your host filesystem read-only & runs comprehensive security scans
🔍 Runs 4 powerful security tools: #Lynis (system audit), #ClamAV (antivirus), #rkhunter & #chkrootkit (rootkit detection)
📋 Logs written to /tmp with structured naming: freshclam, rkhunter_update, lynis, clamav, rkhunter, chkrootkit
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/753/412/467/670/649/original/0dce0098e523cc95.png

2025-12-20 18:55:06 GMT

- reply

📦 Compressed 7z archive created automatically after scan completion
⚙️ Environment variables: HOSTNAME (log suffix), EXIT_AFTER_RUN, CLEAN_LOGS_AFTER_ARCHIVE, CLAM_EXCLUDE_TMP
🔒 Host filesystem mounted at /host in read-only mode for safe scanning
🚀 Available on #DockerHub: michabbb/security-scanner
🔗 https://github.com/michabbb/docker-security-scanner

2025-12-19 22:10:06 GMT

#PhpStorm 2025.3.1 is now available 🚀 🐛
https://youtrack.jetbrains.com/articles/WI-A-231736291
#php #laravel #coding
https://social-cdn.vivaldi.net/system/media_attachments/files/115/748/516/937/313/919/original/aedf093a3475415c.png

2025-12-18 12:01:28 GMT

- reply

📋 #psql compatibility with familiar commands like \dt, \d, \dn, \l
📜 Persistent query history with fuzzy search
🍺 Easy install via #Homebrew, #Cargo or binary download
https://github.com/fcoury/tsql

2025-12-18 12:01:28 GMT

🐘 #tsql — A Modern #PostgreSQL Manager #TUI Built in #Rust #opensource #database #cli #devops
🖥️ Full-screen split-pane interface with query editor and scrollable results grid
⌨️ Vim-style keybindings for modal navigation and editing with customizable config
🎨 Syntax highlighting for #SQL and #JSON powered by tree-sitter
🧠 Schema-aware smart autocomplete for tables, columns, and keywords
✏️ Inline cell editing directly in results grid with automatic SQL generation
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/740/461/352/552/060/original/a4396df317b9eaec.png

2025-12-18 11:09:36 GMT

- reply

☸️ Hardened #Helm Charts for #Kubernetes deployments and new Hardened #MCP Servers for MongoDB, Grafana, #GitHub and more
🛡️ Over 1,000 hardened images and Helm charts already in the catalog - backed by partners like Google, MongoDB, CNCF, Snyk, and JFrog
💼 Enterprise version available: 7-day SLA for critical CVE patches (roadmap: under 1 day), FIPS/FedRAMP/STIG compliance, CIS benchmarks, and custom image modifications

2025-12-18 11:09:36 GMT

- reply

🔍 Full transparency in every layer: Verifiable SBOMs, SLSA Build Level 3 provenance, public CVE data - no hidden or downgraded vulnerabilities
🏗️ Built on trusted foundations: #Alpine and #Debian as proven open-source bases - minimal migration effort, maximum compatibility with existing workflows
🤖 AI-powered migration: Docker's AI assistant automatically scans existing containers and recommends or directly applies matching hardened images

2025-12-18 11:09:36 GMT

- reply

⏰ Extended Lifecycle Support (ELS): Up to 5 additional years of security patches after end-of-life - continuous CVE fixes, updated SBOMs, and audit compliance
🎯 Goal: Secure software supply chain from first line of code to production - secure by default for all 26 million+ developers in the container ecosystem
🌐 https://www.docker.com/blog/docker-hardened-images-for-every-developer/

2025-12-18 11:09:35 GMT

📦 #Docker Hardened Images (#DHI) now free for all developers! #OpenSource #DevSecOps #Security #containers #DevOps
🔐 Minimal, production-ready container images with Apache 2.0 license - fully free to use, share, and build on with no licensing surprises
📉 Up to 95% smaller images with drastically reduced CVEs through distroless runtime approach - smaller attack surface, stronger security
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/740/257/474/706/579/original/e074386f2efdd6ea.png

2025-12-16 23:57:50 GMT

Wouldn't it be cool if your #AI agent could use any #notebooklm notebook for asking questions? Yes, it's super cool 😏 🚀 🕵️ #MCP
🔗 https://github.com/PleasePrompto/notebooklm-mcp
if you are using #claude: there is currently a bug, but till the PR gets merged, you can use a working version from here:
🔗 https://bit.ly/45cCkjc
https://social-cdn.vivaldi.net/system/media_attachments/files/115/731/953/654/673/399/original/286e5b949807a078.png

2025-12-15 14:57:32 GMT

yes, it´s worth the money... #php #phpstorm #laravel #coding
https://social-cdn.vivaldi.net/system/media_attachments/files/115/724/166/909/507/333/original/95743f02b7c92dd8.png

2025-12-15 01:05:42 GMT

Sure, you can use tons of #MCP to create a "council of experts" or you can use https://bit.ly/3YtqbTk - but when your main #ai #coding #cli might be #claude - you can create a skill that uses #codex and #opencode to talk to other #LLM, it’s so easy but powerful at the same time ❤️ 🚀
https://social-cdn.vivaldi.net/system/media_attachments/files/115/720/895/789/673/525/original/7705dd215b6685cc.png
https://social-cdn.vivaldi.net/system/media_attachments/files/115/720/895/881/405/078/original/0fb02a46cac5890f.png

2025-12-14 23:59:08 GMT

⚠️ 🚨 Warning: do not use @FactoryAI ‼️
They are unable to handle any support.
Do not get fooled by their paid YouTube influencers.
Don't expect any help if something does not work.
Here's another "support from hell" story:
https://macropage.medium.com/support-from-hell-or-how-i-call-it-talking-with-factory-ai-droid-e0b255b83373
#ai #coding
https://social-cdn.vivaldi.net/system/media_attachments/files/115/720/634/273/426/694/original/e1937a97c2d8d6ea.png

2025-12-14 00:44:23 GMT

- reply

👨‍💻 Created by Sachin Beniwal, open for feedback and contributions
📚 https://benodiwal.github.io/pg_ai_query
💻 https://github.com/benodiwal/pg_ai_query
https://www.postgresql.org/about/news/pg_ai_query-v010-first-stable-release-with-multi-model-ai-for-postgresql-3191/

2025-12-14 00:44:23 GMT

- reply

🔍 Schema-aware query intelligence through secure introspection understands your database structure for accurate query generation
🔌 Multi-model flexibility: supports #OpenAI, #Anthropic, #Gemini, #OpenRouter compatible APIs, and fully local inference via #Ollama
🖥️ Native support for Linux and macOS, compatible with #PostgreSQL 14 and newer versions
🛠️ Keeps your entire AI-powered SQL workflow inside Postgres — no external tools required

2025-12-14 00:44:23 GMT

#pgaiquery v0.1.0 — First stable release brings multi-model #AI directly into #PostgreSQL #opensource #database #SQL
🤖 Generate SQL from natural language queries right inside Postgres: `SELECT generate_query('list customers who have not placed an order in the last 90 days')`
📊 AI-interpreted EXPLAIN ANALYZE translates query performance data into plain language explanations
⚡ Automated index and rewrite recommendations help optimize database queries without manual analysis
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/715/149/853/317/543/original/37be5cd9a7787e4e.png

2025-12-13 23:46:59 GMT

- reply

📋 Global configuration in ~/.mdflow/config.yaml for default frontmatter per command, logs written to ~/.mdflow/logs/ for debugging, LiquidJS template engine supports conditionals, loops and filters
https://github.com/johnlindquist/mdflow

2025-12-13 23:46:58 GMT

- reply

📁 Build your own agent library: Create ~/agents/ directory with specialized agents like review.claude .md, commit.gemini .md, explain.claude .md, add to PATH and use from anywhere
🛠️ Installation via npm install -g mdflow or bun install && bun link, run mdflow --setup for shell integration to make .md files directly executable

2025-12-13 23:46:58 GMT

- reply

📥 Powerful import system directly in prompts: @./src/api.ts for single files, @./src/**/*.ts for glob patterns, @./file.ts:10-50 for line ranges, @./types.ts#UserInterface for TypeScript symbol extraction, @https://url for remote content
🎯 Template variables for dynamic prompts: {{ _stdin }} captures piped input, {{ _1 }} {{ _2 }} for positional arguments, {{ _args }} for numbered list of all args, custom variables like {{ _feature_name }} with CLI overrides --_feature_name "value"

2025-12-13 23:46:58 GMT

- reply

⚙️ Print mode runs by default for non-interactive execution, interactive mode enabled via .i. filename marker (task.i.claude .md), _interactive frontmatter key, or -_i CLI flag
🌐 Remote URL imports cached at ~/.mdflow/cache/ with 1-hour TTL, use --no-cache to force fresh fetch, automatic .env file loading from markdown file directory with support for .env.local, .env.development, .env.production

2025-12-13 23:46:57 GMT

📝 #mdflow - Turn #Markdown files into executable #AI agents! Multi-backend CLI for running .md files against #Claude, #Codex, #Gemini or #Copilot #opensource #AI #DevTools
⚡ Filename determines command automatically: task.claude .md runs with Claude, task.gemini .md runs with Gemini, task.codex .md runs with Codex, task.copilot .md runs with Copilot
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/714/923/867/178/757/original/714c060fa0316b6f.png

2025-12-13 23:46:57 GMT

- reply

🔧 YAML frontmatter converts directly to CLI flags: model: opus becomes --model opus, mcp-config: ./mcp.json becomes --mcp-config ./mcp.json, dangerously-skip-permissions: true becomes --dangerously-skip-permissions
🔄 Full Unix philosophy support: Pipe data in and out with stdin/stdout, chain agents together like git diff | review.claude .md or plan.claude .md | implement.codex .md, transparent logging shows what runs

2025-12-13 22:39:45 GMT

- reply

🔔 Built-in alerting support via #vmalert integration. Fits RUM (Real User Monitoring) and SIEM use cases. Automatically parses CEF messages sent over Syslog protocol.
🖥️ Built-in web UI for log exploration. #Grafana plugin available for building arbitrary dashboards. Interactive command-line tool vlogscli for querying.
📊 Benchmark results show VictoriaLogs outperforming Elasticsearch, MongoDB, DuckDB, PostgreSQL in JSONBench and ClickBench comparisons.

2025-12-13 22:39:45 GMT

- reply

☁️ Available as Docker images on Docker Hub and Quay, binary releases on GitHub, and Helm charts for #Kubernetes deployments. #Apache2 license - open source and free for commercial use.
🌐 https://github.com/VictoriaMetrics/VictoriaLogs

2025-12-13 22:39:44 GMT

- reply

📈 #LogsQL query language provides fast full-text search, advanced analytics, and data extraction/transformation at query time. Can be combined with Unix tools like grep, less, sort, and jq for log analysis.
🎯 Optimized for high cardinality fields like trace_id, user_id, and ip addresses. Supports logs with hundreds of fields (wide events), multitenancy, out-of-order ingestion, live tailing, and surrounding log context selection.

2025-12-13 22:39:44 GMT

- reply

🔧 Zero-config single executable setup - no tuning required for VictoriaLogs or the operating system. Reasonable defaults automatically adjust for available CPU and RAM resources. Much easier to set up and operate compared to Elasticsearch and Grafana Loki.
🔌 Accepts logs from popular collectors including Fluentbit, Fluentd, Logstash, Vector, Promtail, #OpenTelemetry, DataDog Agent, Filebeat, Telegraf, Journald, and Syslog. Native protocol support for seamless integration.

2025-12-13 22:39:44 GMT

- reply

⚡ Uses up to 30x less RAM and 15x less disk space than Elasticsearch and Grafana Loki. Capacity and performance scale linearly with available resources - runs smoothly on Raspberry Pi and on servers with hundreds of CPU cores and terabytes of RAM. Scales horizontally to hundreds of nodes in cluster mode.

2025-12-13 22:39:43 GMT

📊 #VictoriaLogs - Fast #opensource log database from #VictoriaMetrics using 30x less RAM & 15x less disk than #Elasticsearch and #Loki. Handles terabytes of logs efficiently with the powerful #LogsQL query language. Available in single-node and cluster mode. #observability #devops #logging
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/714/659/551/435/045/original/1991040940aaff7a.png

2025-12-11 22:03:21 GMT

- reply

🔐 AES encryption & space-saving compression protect your data from unauthorized access
💿 Creates rescue disk & backup images in VHD/VHDX format - opens
directly in #Windows Explorer
🛡️ Protects against data loss from viruses, hackers or defective hard drives - suitable for beginners
🌐 https://www.oo-software.com/de/special/pdi912
deal source: https://www.mydealz.de/share-deal/2692221

2025-12-11 22:03:20 GMT

📦 #OODiskImage 21 Premium - #Free full version in #ComputerBILD Advent Calendar! #Backup #Windows #DataProtection #devops
💾 Flexible backup solution for #Windows11 & #Windows10: Backs up drives, partitions, directories & individual files
⚡ 1-Click Backup: Saves complete PC including operating system & all connected drives with one click
🔄 Supports regular, incremental & differential backups for optimal data security
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/703/191/873/024/640/original/771ed37afc8ca448.png

2025-12-11 10:25:14 GMT

- reply

⚙️ Configuration: `cache.result` for result caching, `extensions` for custom listeners
🐛 Bug fixes: Parallel mode stability, Windows path handling, PHAR extraction issues
https://github.com/sebastianbergmann/phpunit

2025-12-11 10:25:13 GMT

#PHPUnit 12.5 released - key features:
🧪 Enhanced parallel testing: Improved worker coordination and resource management for faster test suites
#PHP #Testing #PHPDeveloper #TDD
📊 Better failure output: Inline diffs for arrays/objects, clearer stack traces with context
📈 Code coverage: New metrics for line/branch execution, HTML reports with filters
🔍 Mock assertions: `assertMockObjectCreated()` and `assertMockObjectNeverCreated()`
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/700/446/673/586/775/original/753b49e0ab510aa8.png

2025-12-11 08:56:45 GMT

- reply

🌐 https://mistral.ai/news/devstral-2-vibe-cli

2025-12-11 08:56:44 GMT

🚀 #Devstral2 & #MistralVibe CLI: State-of-the-art #opensource agentic coding models #AI #coding #developers
🔥 #Devstral2 (123B) achieves 72.2% on SWE-bench Verified — SOTA for open-weight code agents under modified #MIT license
⚡ #DevstralSmall2 (24B) scores 68.0% on SWE-bench, runs locally on consumer hardware under #Apache2 license
💰 Up to 7x more cost-efficient than Claude Sonnet at real-world tasks
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/700/098/876/355/705/original/7e04538f240d9658.png
https://social-cdn.vivaldi.net/system/media_attachments/files/115/700/098/922/000/759/original/569613be541743a9.png

2025-12-11 08:56:44 GMT

- reply

🧠 256K context window supports multi-file orchestration with architecture-level reasoning
🖥️ #VibeCLI: #opensource terminal agent with project-aware context, smart file references & Git integration
🔧 Compatible with #Cline #KiloCode & available as #Zed extension for IDE integration
📊 Pricing: $0.40/$2.00 per million tokens (Devstral 2), $0.10/$0.30 (Small 2) — currently FREE via API
🖥️ Devstral 2 requires 4x H100 GPUs, Small 2 runs on single GPU or CPU-only configs

2025-12-10 16:04:40 GMT

- reply

@nprofile…zmtt glad i could help, best of all, it was free.... next time, it's not... oh i forgot... you won´t ask me anything in the future, i´m so so so sad about that, i guess i need to cry 3 hours alone in the dark now.... yeah, i guess that will help to process this drama......
buhhhhhhhhhhhhh.... 😭 😭 😭
https://social-cdn.vivaldi.net/system/media_attachments/files/115/696/095/129/399/118/original/c03b79516df7fa7b.mp4

2025-12-10 09:59:48 GMT

- reply

@nprofile…zmtt people who don't understand docker, yes, are too lazy or too dumb, there is no other logical explanation, if that sounds arrogant, I don't care... It's not my mission to teach others, if others work like in the stone age, they should.... I'm very sick having this conversation, sorry, nothing personal....

2025-12-10 01:03:54 GMT

- reply

🔧 AI-powered features: automatic search query and strategy configuration, real-time updates via #Supabase Realtime subscriptions, AI-generated one-sentence summaries with semantic embeddings for each successful execution, manual "Run Now" button for immediate execution without waiting for cron
🔐 Authentication options: #Supabase Auth with email/password and #Google OAuth support, Row Level Security (RLS) policies ensure users only access their own scouts, messages, and executions

2025-12-10 01:03:54 GMT

- reply

🏗️ Scalable dispatcher architecture designed for thousands of scouts: pg_cron runs every minute to dispatch due scouts, pg_net fires individual HTTP POST requests, each scout runs in isolated #Deno edge function
with full resources (256MB memory, 400s timeout), automatic cleanup job removes stuck executions every 5 minutes, credentials securely stored in Supabase vault

2025-12-10 01:03:53 GMT

🔍 #OpenScouts: Create #AI scouts that continuously search the web and notify you when they find what you're looking for #opensource #NextJS #React #TypeScript #Supabase #OpenAI #Firecrawl
⚡ Built with cutting-edge tech stack: #NextJS 16 with App Router & Turbopack, #React 19, #TypeScript, #TailwindCSS v4, #Supabase for database, auth & edge functions, #pgvector for vector embeddings and semantic search, #OpenAI API for AI agent & embeddings, and Resend for email notifications
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/692/577/157/735/813/original/b9c7aa954a343f7c.png

2025-12-10 01:03:53 GMT

- reply

🎯 Create automated scouts that run on customizable schedules (hourly, every 3 days, weekly) to continuously monitor the web - whether tracking AI news, finding new restaurants near you, or any other updates you want to follow, scouts work 24/7 to find what you need and notify you when they discover it