What actually shipped, and how fast
Over the past ten days, hashtag.org's changelog shows a steady cadence of releases rather than one big launch: five separate updates dated July 12 through July 21, ranging from a single-fix day (the July 17 WebGL health check) to a 26-commit release on July 20. The July 12 entry alone covered nine new features and six bug fixes tied to the RADIO_TTS_PROVIDER work and MCP architecture improvements, logged as a 20-hour day.
The pattern across entries is consistent: a mix of feature work and fixes shipped together, with day-of writeups describing hour counts and commit totals for each release. July 19 logged 12 features and 6 fixes over 23 hours; July 13 logged 13 commits (6 features, 7 fixes) over about 9 hours; July 18 logged 8 changes over roughly 12 hours.
This is a self-reported build log, not an audited engineering report, but it's the record we're putting on the public record for anyone tracking how the platform evolves week to week.
In plain terms
This section is just the receipts: what got built, on which days, and how long it took, straight from our own public changelog. We're not polishing the numbers, just reporting what we logged.
Discovery infrastructure: super-map, keyword scanning, MCP
The July 21 release focused on the super-map, the geo-spatial layer that lets agents query portals by location and intent rather than crawling pages. Shipped features included enhanced super-map functionality, expanded keyword scanning, in-app reporting configuration, and an admin control room setup, alongside one critical fix, over what the changelog describes as a 14-hour day.
Keyword scanning matters because it's the mechanism agents use to filter portals by hashtag keyword (#coffee, #plumber, #lawyer) rather than relying on GPS proximity alone, per the Agentic Internet documentation. Expanding that scanning is a discovery-layer change, not a cosmetic one: it affects what an agent querying the map actually gets back.
The July 12 release also touched core plumbing for agent connectivity, shipping MCP enhancements alongside a new RADIO_TTS_PROVIDER integration. The MCP server is what lets an external agent, whether that's a Claude instance, a Cursor session, or a browser agent, connect to the hashtag network and search it, leave leads, buy, book, or claim its own #name presence, according to the platform brief.
July 13's release added compute node humanization, SEO security fixes, and new compute marketplace features, extending the builder-facing side of the platform (six feature commits, seven fixes).
In plain terms
This is the part that lets AI agents actually find and use businesses on the platform, by location or by keyword, and the part that lets other AI tools (like Claude or Cursor) plug straight into the network to search it or book things. Think of it as the roads and doors agents use to get in.
GIGI avatars, onboarding, and the creator-facing funnel
July 20 was the heaviest release of the window: 26 commits over a 21-hour day, split into 11 features/improvements and 15 fixes, centered on the GIGI avatar experience and the marketing funnel for creators. Shipped items included a GIGI avatar cache-bust, a marketing funnel update, footer improvements, and an admin analytics fix.
GIGI is the per-portal AI agent that answers visitors in chat, voice, and video (using Tavus personas per the platform brief), takes bookings, and captures leads for the portal owner. A cache-busting fix for the avatar suggests the team was chasing a stale-asset bug rather than adding new capability, but it sits alongside real feature work on the creator funnel.
The July 19 release built on that with onboarding and analytics work: an onboarding dropdown added to the homepage, homepage UI fixes, deeper analytics integration into the admin dashboard, and an improved humanizer component, over 12 features and 6 fixes across a 23-hour day. July 18 followed with a new analytics command center and enhanced session replay, plus fixes to OAuth issues and GIGI bugs.
In plain terms
GIGI is the AI 'employee' each business portal can run, answering questions and taking bookings on video or chat. This stretch of work was mostly about making that avatar reliable and making it easier for new creators to sign up and see their own analytics.
Newsroom, app store, and reliability fixes
The July 16 release launched two structural additions: a weekly agentic-web newsroom, now live at /news with an article template and cadence, and a centralized app store, alongside performance and SEO enhancements, logged as a 17-commit, 18-hour day. This newsroom is the publication you're reading now.
The following day's release was narrower but targeted at a real user-facing bug: users with broken WebGL were seeing a black map instead of the spatial globe. The July 17 fix added a WebGL health check and a raster fallback so those users get a working, if less fancy, map instead of nothing.
July 15 shipped a tube channel video strip integration plus schema improvements and authentication fixes, eight changes total. July 14 was the lightest release of the window: footer optimization, orphaned-page linking, and a downloads page update, three commits in about an hour.
In plain terms
This part covers the newsroom you're reading, a new app store, and a handful of fixes so the site actually works for everyone, including a fallback for people whose browsers can't render the interactive map properly.
Where this fits against the rest of the agentic web
The company profiles in this same newsroom describe a landscape dominated by labs and platforms sending agents outward: OpenAI's Operator handles consumer task automation and e-commerce, Anthropic's Claude computer-use work targets enterprise productivity and document workflows, Google DeepMind's Project Mariner does consumer browsing and research, and Perplexity AI pairs agentic search with automated browsing and shopping. These are, broadly, agents going out to use the existing web.
hashtag.org's own profile in that same landscape report sits under 'Agentic Web Infrastructure,' described as the receiving end where agents land, read, and transact. The releases covered here map onto that role directly: the super-map and keyword scanning make portals findable by location and intent, the MCP server lets any outside agent (including tools built by the labs above) connect and act, and GIGI gives each portal its own agent to answer, book, and capture leads when one of those outside agents (or a human) arrives.
The platform's discovery layer also works differently from a ranked search index. Keyword ranking on hashtag.space runs as an open, on-chain auction using $SPACE, so who gets found is a visible stake rather than a hidden algorithm, and reputation and portal identity are meant to be verifiable on-chain rather than sitting in one company's database. hashtag.org's traffic data for the last 30 days already shows more bot and AI fetches (4.1K) than human visits (3.6K), a rough signal that agents are already the audience this infrastructure is built to serve.
None of this competes head-on with what OpenAI, Anthropic, Google DeepMind, or Perplexity are building: they're improving the agents themselves. hashtag.org is building the side those agents need to land on: structured, agent-readable, and, per the platform's model, funded by returning proceeds to users in web3 rather than outside investment.
- OpenAI (Operator), Anthropic (Claude computer use), Google DeepMind (Project Mariner), Perplexity AI: agents that go out and browse, compare, and transact on the existing web.
- hashtag.org: #names, geo-pinned portals, GIGI per-portal agents, an MCP server, and open on-chain keyword auctions, the infrastructure that makes a business findable, readable, and transactable when one of those outbound agents arrives.