... format. it captures a title and a list. it does not capture which install, which patient, which deploy. and it does not capture who signed off on it. no signer means no real credential. a real credential has four rows. the work, one sentence in plain language. the artifact, the photo or the link. the peer signature, one verified colleague. the customer signature, the actual buyer or user. the cost-of-credential argument matters here. a linkedin endorsement costs ... ... signer nothing. so it signals nothing. a tam signature costs the signer their public reputation if the work was a lie. so it signals everything. cheap credentials are worthless credentials. this is what we shipped at tam network. four rows

... refresh. Instead of blocking on a fresh query and leaving you looking at a loading state, Basedash shows the most recent results right away and kicks off the refresh in the background, swapping in new rows the moment they re ready. The refresh itself moved off the web request and onto a dedicated background job, with duplicate refreshes for the same query collapsed into one. The chart header keeps a subtle spinner (with an Updating data from tooltip) while ... ... dashboard ever going blank on you. A cleaner, smarter AI chat The AI chat s thinking steps got a visual refresh. Reasoning summaries now render inline as plain text like preambles instead of hiding behind an expandable Reasoning summary row

... great but Python-only for everything). So I built ApiTap: a managed extraction engine that runs on a 256 MB worker. Rust + Apache DataFusion under the hood. Every number on apitap.dev is measured, not projected: 1 BILLION rows in one endurance run (99.8% success) 3.7x faster than Airbyte on identical 1.5M-row

... cross-organization queries stay locked down. Fixes and improvements Restored file-based Fivetran connectors (like S3 and SFTP) so they create and sync correctly again. Fixed chart images in insight emails and Slack so they render with the correct row-level-security context instead of showing the wrong data. Made Connection issues sidebar rows

... exactly that: shipped the entire result set to the browser and let it sort itself out. Same deal for Spanner. A lot of Basedash is just "run the SQL the user wrote, show the rows." We support 10+ dialects (Postgres, MySQL, BigQuery, Snowflake, Athena, SQL Server, Spanner, and friends), and every query gets paginated because we can't stream a 12M row

... additions on the Repos tab that turn it from a passive listing into a dashboard you actually live in. CI status, at a glance Every Repos row now leads with a coloured dot sourced from the default branch's status-check rollup: - green passing - red failing - yellow pending / expected - dim no rollup (no CI on this repo) Press s and the sort cycle picks up a "by CI" entry that surfaces failing repos first open octoscope and immediately ... ... rows

... built agency-os . It turns Notion into a single board for working with AI: You talk through ideas with the agent. It pushes back, asks clarifying questions, and scopes the work into tasks and subtasks on the row. You approve before anything runs. Nothing dispatches autonomously - every execution is operator-gated. Agents pick up approved rows

... drill-in pattern I prototyped for Repos in v0.10.0 now works on PRs and Issues too and it's bound to Enter , the universal TUI gesture for "go deeper". What you get when you press Enter on a row: PRs state chip (Open / Draft / Merged / Closed), markdown-rendered description, reviewers (requested + completed), checks summary, files changed, recent commits, labels, curated timeline. Issues state chip, markdown-rendered description, assignees, recent comments, labels, linked PRs (the ones that would close the issue ... ... zero rate-limit pressure. Descriptions and comments render through glamour ( https://github.com/charmbracelet... ) with chroma syntax highlighting, so a "## heading" is a heading and a fenced code block is colourised. Heads up breaking UX change: Enter on a list row

... that gives you a clean, professional sidebar to manage your spreadsheet data without ever touching a formula or navigating complex menus. The problem I was solving: Excel is powerful, but managing data in it feels clunky. Adding rows means scrolling to the bottom, editing means finding the right cell, validating means writing formulas. It shouldn't be this hard. What GridManager does: Add, edit and delete rows

... each error. Error Type Average Estimated Cost Pricing error (showing higher than actual) 47,000 Pricing error (showing lower than actual) 23,000 Missing critical feature 38,000 Wrong market positioning 52,000 Outdated company narrative 31,000 Competitor credited for your work 44,000 Total estimated losses across 50 founders: 1,204,000 Average per founder: 28,000 Per error average: 18,500 These are not hypotheticals. These are founders calculating actual lost deals, longer sales cycles, and confused ... ... told us 'we checked and you're not SOC2 yet.' I couldn't argue they trusted AI over us." Estimated loss: 180,000 Founder #27 (Dev tool, 1.2M ARR) "Perplexity claimed our free tier had a 1,000-row

... convenient. They let you ship faster and keep the interface tidy. I kept wanting to add kebab menus and quick popups because they re familiar and compact. But those patterns encouraged tuck it away for now decisions. In a row-based workspace, a kebab menu expands outward and detaches the controls from the row

... Hindi, Spanish, and Thai and get results. Supports Major Databases: Currently works with SQL Server, MySQL, and PostgreSQL . Other database types coming soon. Large Database Friendly: Designed to scale whether your database has thousands, millions, or even billions of rows, DataQueryAI loads results efficiently with paging and schema guidance. End-to-End Analysis: Query visualize export reports. Export Options: Export results in HTML, CSV, or XLSX formats. Privacy-Focused: Option for local execution ensures your sensitive data stays secure. Designed ... ... Should Try It: Analysts who want faster insights without writing SQL Product managers and founders seeking ad-hoc data analysis Small teams or startups that need lightweight, powerful analytics Users working with huge datasets , from millions to billions of rows

... spot the next trend before it blows up. Comment from Maksim Ryndin(@maksim_ryndin): Front-end: mostly @Gemini , a bit of @Claude Code . In browser chats, prefer avoiding to give any access to the operating system. Backend: with @VS Code , usually with Rust (even for browser extensions like https://semplico.app/sleach ), mostly myself Comment from Tom Eldridge(@teldridgeldn): Running Claude Code inside Cursor and as a VS