eummm

looks cool, would be great to try effects with some simple guitar riff as a sample to try before plugging a guitar into a mic

very cool! thanks for making it oss! consider adding few short audio examples on github's Readme (or just as video/yt), it would be great

Hi HN! Made this AICW Video open-source tool when had to quickly edit a number of raw video interviews.

It can: - find which separate audio file belongs to which video and replace it (use case: audio recorded on a separate phone) - auto crop a video (manual crop too) - suggests clips to create - adds dynamic captions - exports to TikTok/Insta/others - can blur a face, can overlay audio with tts

Uses local AI for audio to text, face detection. For scenes detection requires Claude or Codex but can use Ollama with Qwen. Mac only. Works standalone and as MCP via Claude/Codex.

AI is eating SaaS

The purpose of this widget:

- show summary of the current page (takes it from meta description tag);

- allow user to ask their favorite AI (ChatGPT, Claude and others) to summarize current page (why not? it is convenient!)

- share website via messengers / social networks similar to how AddThis worked

Website and live demo: aicw.io/summarize-widget

Source-code: github.com/aicw-io/aicw-summarize

added monkey patch that works with 8.1.2 https://github.com/rails/rails/pull/57128#:~:text=Update%202...

thank you for feedback! will think how to improve

adding text to speech is a great idea, thank you!

HM looks great, but looks like they collect lot of personal data.

From their privacy policy:

The personal information we collect may include:

Name User name Email address Gender Birthday (month/year) Ethnicity Relationship Status Number of Family Members/Children/Children under 13 Experience with Meditation User generated content We may collect other information that does not reveal your specific identity, such as:

IP Address or other unique device identifier Information collected through cookies, pixel tags or other technologies App usage data Geo-location information User generated content Device generated data

SQLite3 is awesome but turned out if you create a new RoR application with table and child table with child records and then rename or remove a column then all child records are deleted. This issue is currently present in both 8.1.3 release and edge versions.

To grade AI visibility of your website one can use this command line:

npx @aichatwatch/aicw ai-visibility EXAMPLE.com

It checks how the server responds to AI crawler bots, analyzes the content (HTML) code, checks if the website is present in public datasets, and returns a final grade similar to this:

================================================================================ AI VISIBILITY REPORT for https://aichatwatch.com/ on Oct 26, 2025, 23:11 PM ================================================================================ SCORE: 86% (93.5/109 points) GRADE: B ================================================================================

DETAILS: [PASS] Server: check /robots.txt.................... 15/15 [WARN] Server: check /sitemap.xml................... 0/5 [PASS] Server: X-Robots-Tag HTTP Header Check....... 15/15 [PASS] Server Response Speed........................ 6.3/7 [WARN] Content: JSON-LD Structured Data Presence.... 0/5 [PASS] Content: Blocking Meta Tags Check............ 8/8 [PASS] Content: Server-Side Rendering (SSR) Check... 8/8 [PASS] Content: HTML code structure................. 4.2/5 [PASS] Content: Mobile Version Availability......... 7/7 [WARN] Dataset: /llms.txt available................. 0/1 [PASS] AI bots access: Data for Training............ 12/12 [PASS] AI bots access: Search Indexes............... 10/10 [PASS] AI bots access: User Interactions............ 8/8 [WARN] Dataset: Common Crawl Dataset................ 0/3

-------------------------------------------------------------------------------- ISSUES: -------------------------------------------------------------------------------- Server: check /sitemap.xml Impact: 5 points lost Details: No /sitemap.xml found

    Content: JSON-LD Structured Data Presence
     Impact: 5 points lost
     Details:  No JSON-LD structured data found.

    Dataset: /llms.txt available
     Impact: 1 point lost
     Details:  No /llms.txt found (not critical)

    Dataset: Common Crawl Dataset
     Impact: 3 points lost
     Details:  Not found in 3 recent crawls: CC-MAIN-2025-43, CC-MAIN-2025-38, CC-MAIN-2025-33

The book about AI search engines and how to optimize for them. The e-book (for Kindle) is free until October 24, 2025, 11:59 PM PDT

There are some demo reports tracking who and how its mentioned on aichatwatch dot com, like: "Y Combinator", "Top AI Tools", "high blood pressure and hypertension", "Best Cities for Nomads"..

Hey HN! creator is here. AI Chat Watch is an open-source CLI tool that helps marketers to track and understand how AI chatbots talk about brands, products, industry.

It runs a set of questions across multiple AI models (currently ChatGPT, Claude, and Perplexity via OpenRouter) and generates reports showing: - Which entities are mentioned, how often and in which order - Trends of these mentions (comparing to previous dates) - Specific links and domains cited in responses by AIs

Originally built this as a commercial product for marketers doing competitive research, but decided to open source it so anyone can run it locally. You'll need to provide your own OpenRouter API key (~$0.30 per run for 5 questions × 3 models).

Some example use cases: - Competitive analysis (e.g., "best CRM software") - Keywords monitoring across AI responses - Exploring websites and specific pages cited by AI - SEO/visibility in AI-generated answers

Installation is straightforward via terminal. There are demo reports in the repo to see what the output looks like.

Would love feedback and suggestions on making it more valuable Thanks

Thank you for your video feedback, subscribed to your channel! btw really enjoyed the background music btw!

glad you found value in the report and your point about hallucinations, we'll try to implement additional fact-checking measures.