New — Works without an SEO plugin. With no SEO plugin active, Agentimus now covers the search basics itself: per-page SEO titles, social share cards (Open Graph, with a default share image you pick in Settings), canonical links, and a sitemap that carries the last-changed dates core’s own leaves out. With Yoast, Rank Math, SEOPress, AIOSEO or The SEO Framework active, nothing changes — it detects the plugin, steps aside on every overlapping surface, and the readiness report names who is doing what. The switch is automatic, in both directions, the moment an SEO plugin is activated or deactivated.
New — The weekly email arrives when you say. Settings → Weekly email gained a day-and-time picker — the note lands on the day and hour you choose, your site’s clock, instead of always Monday morning.
Improved — Acting on the review queue reads back instantly. Blocking or allowing a client now updates the admin menu’s count bubble the moment you click, not up to a minute later; cards that offer no one-click Allow/Block say why (a real search engine’s name can’t be safely blocked by name); and the dashboard’s downward trend arrows dropped the error-red for a calm amber — a quieter week is information, not a fault.
1.28.0
New — A weekly email about your site’s AI week. Once a week, Agentimus emails the site owner a short note: how often AI agents read the site against the week before, readers arriving from AI answers, impostors caught, the readiness score and its move since the last note, and one thing worth doing. Built only from data already stored on your site and sent with WordPress’s own mail — no images, no tracking, nothing leaves your server except the email itself. A week with nothing to report sends nothing, and every email ends with a one-click stop link.
New — The review queue shows itself. The count of clients awaiting your review now rides the Agentimus entry in the admin menu — the same red bubble core uses — visible from every admin screen and kept current while you work. When the queue empties, it disappears.
1.27.1
Fixed — The review-ask endpoint answers a permission denial before a validation complaint, like every other route. No security impact — the capability gate always held; only the order of the two refusals was wrong.
1.27.0
New — One honest review ask. After Agentimus has been on your site a week, actually been used, and your readiness score is healthy, one quiet Dashboard card asks whether it has earned a WordPress.org review. Any answer puts it away — “Maybe later” for a month, the others for good — and once answered, the admin footer’s rating line retires too. Never a site-wide notice, no incentives, nothing sent anywhere.
New — Setup ends with proof. A new install’s finishing screen now shows what just went live — llms.txt and the discovery document as clickable addresses, already carrying the words you typed a moment earlier — plus the site’s starting readiness score. Empty dashboard cards now say what will appear in them, and roughly when, instead of a bare zero.
Improved — The MCP server card describes its authentication in standard terms — http/basic and oauth2 — so the card and a connecting agent’s documentation finally use the same words.
1.26.0
Improved — Reading ease now measures honestly: sentences end at block boundaries (bullet lists stop grading as run-ons), a page’s own recurring subject terms count as familiar words instead of being charged every use, and code samples are excluded from prose grading (a mostly-code page skips the grade rather than mis-scoring). A pass names the terms it forgave; a warn names the page’s heaviest words.
New — Agent writes follow the readability rules: the MCP write tools show connected agents the same drafting bars the in-admin assistant writes to, and every create/update answers with the post’s readability grade so the agent can fix its own draft.
Fixed — The Optimize worklist names real content types (“3 Posts, 1 Page” — not everything a “page”), and every readability check has its own plain guidance line.
1.25.0
New — A writing assistant inside wp-admin. A quill button opens a drawer on every Agentimus screen: describe the post you want, edit the outline it proposes, and preview the complete draft — real editor blocks, AI description, topics, suggested categories and tags — before anything is saved. Create draft opens the post in the editor; the assistant never publishes. It needs the agent-writes switch and a WordPress AI provider (7.0+), and your Content Guidelines steer its voice automatically. It also writes to the same bar the AI Readability panel grades — a liftable opening summary, quotable paragraph lengths, plain sentences, cited sources — so fresh drafts arrive with their checks green.
New — Long articles, written in parallel. Once you approve an outline, the assistant writes every part of the article at the same time — the introduction, each section, the closing and the title details — and assembles them in your outline’s order. Each section gets its own budget, so an article is no longer capped by what fits in one response; progress fills in on the outline as parts land; and a part that fails — a rate limit, a network hiccup — retries alone instead of rerolling the whole article. Even reloading the page mid-write resumes where it left off.
New — It edits existing posts too. Pick a post, describe the change, and review the revision before applying it. A post’s status is never touched — published stays published — WordPress revisions keep every prior version, and unusual block layouts are declined honestly instead of mangled.
New — Images live where you write. Drafts arrive with alt-filled image placeholders; every image block gets Generate image from the alt text, and a Featured image (AI) sidebar panel drafts a hero from the title — or pick from your library as always.
New — Ask AI on any block. Select a paragraph, heading, list or quote and say what to change — rewrite, shorten, or “add a conclusion after this”. Unchanged text is preserved word for word, and normal undo brings anything back.
New — A featured-image check, with the fix one click away. AI Readability now checks that a post has its featured image — the picture link previews and embeds show — and skips honestly where a content type offers none. In the block editor the warning carries Generate with AI: the same title-seeded featured-image generation, offered exactly where the gap is reported.
New — Every dashboard number is a door. The By-endpoint and Top-clients rows open the Request Log pre-filtered to that endpoint or client, and the Traffic-from-AI sources and landing pages open the AI Traffic report the same way. Hovering any row shows the numbers behind it: a trend arrow reveals the two half-window counts it was computed from, a referral row its share of the window — no arrow or percentage asks to be taken on faith.
Improved — One Agentimus box in the editor. The AI description and Topics panels merged into the branded Agentimus box, and AI Readability gained two checks: cited sources and reading ease.
Improved — The admin wears its marks. Every navigation tab and dashboard tile carries a small icon in the house style — the Discovery-listed APIs join the tiles as a fourth count — and the masthead, gold-gem arrow included, is a real link back to the plugin page. Report screens now read as card stacks: AI Traffic splits into overview, top sources & landing pages, and a by-day timeline; the dashboard’s endpoint and client lists become cards of their own with plain-language leads; scanner blocking hands its bot-identity half a card of its own; and the Advanced tab’s developer & maintenance zone sits behind a labelled rule instead of an amber tint.
Improved — The assistant’s drawer is easier to leave and harder to mis-click. A ← Back beside the Close walks you up a level, the revise bar lives in the pinned foot so it’s always in reach, editing a draft says Update draft while a published post keeps its stays-published warning, and every consequential act — revising, updating, creating, drafting over held work — states its consequence in a small confirm first. The drawer is a little wider, and the post picker explains its list: your ten most recently edited, with search reaching every post.
Improved — Charts you can tap on a phone. The daily bars keep a touch-friendly width and scroll sideways, opening at the newest days; on the AI-traffic chart even a quiet day is a full-height tap target now.
Improved — AI errors in plain language. A quota wall now names the fix (pick from the library, or check the provider’s plan), a provider error that arrives without details still becomes a human sentence, and no AI failure returns a bare 502 — a status some CDNs replace with their own error page.
Improved — The Cited rung updates itself. The AEO/GEO score rail refreshes the moment an AI Visibility check finishes or its results are cleared — no more waiting for the next visit to notice.
Fixed — Agent field edits refresh the public files. When a connected agent set a page’s AI description or topics over MCP (a meta-only write), the database changed but cached public surfaces — the page’s .md twin, llms.txt — could keep serving the old text until the next real save. Those writes now fire the same cache flush a normal save does.
1.24.0
New — Verify more bots — and edit the list yourself. Verification used to cover the five search engines that publish reverse DNS. Agentimus now also checks a bot against the IP ranges its operator officially publishes, which adds GPTBot, OAI-SearchBot and PerplexityBot — and the whole list is yours: Settings shows every verified bot with how it’s checked, any built-in can be switched off, and when a new operator starts publishing verification data you can add it yourself instead of waiting for a plugin update. The range lists refresh once a day in the background, never while serving a visitor. An unreachable publisher just means “not checked”: a match can verify against an older copy, but only a fresh copy may ever call a bot fake.
New — A proven impostor is refused, not just flagged. With blocking and verification both on, a client caught lying about who it is — a fake “GPTBot”, a forged “Googlebot” — is turned away from the AI endpoints with a 403, as part of the auto-deny-spoofed switch (an identity forgery is the same kind of deception, just proven). Your rules still outrank everything: an agent on your Allow list is never denied, a bot on your Block list stays blocked even when it verifies as genuine, and anything unclear — slow DNS, a stale range list — is served, never guessed at.
Improved — The review queue explains its own housekeeping. A client that’s merely new leaves the queue by itself 48 hours after its first visit; that always happened silently, and looked like a bug. Its chip now counts down (“New · leaves in 31h”), and a fixed footer says who leaves on their own, that flagged clients stay until you act, and that every request stays in the log — with a Manage clients link right there. The queue’s title and filter tabs stay put while the list scrolls, and an impostor’s card now says exactly how it was caught (“its address isn’t in GPTBot’s published IP ranges”) and notes when it’s already being refused.
Fixed — The admin fits your phone. No more sideways panning on narrow screens, dialogs size themselves to the real visible area (the Close button can no longer land beyond the screen’s edge), and the About page no longer clips its text.
Changed — screen names are consistently title-cased: AI Traffic, Request Log, Agent Access, About Agentimus.
1.23.0
New — Agent access alerts on every visit. The nav pill now re-lights whenever an agent authenticates or acts — not only the first time ever — and the log lists the newest activity first. The screen also gained its proper page header.
New — Your voice, everywhere AI writes. If your WordPress defines Content Guidelines (the experimental brand-voice and copy-rules feature), Agentimus honors them automatically: Draft, Suggest and Fix with AI follow them, and AI agents connected over MCP see them right on the draft/edit tools — so agents write in your voice too. Nothing to configure; your guidelines never appear on any public endpoint.
Improved — Markdown twins ready for WordPress 7.1. The new Tabs block converts with each tab clearly labelled, Playlists become a readable track list, tables come out as real Markdown tables instead of run-together text, and preformatted blocks keep their line breaks — diagrams built in the visual editor no longer flatten to one line.
Fixed — Full description on WordPress.org. The plugin directory folds extra readme sections into the description before applying its word limit, so the listing was still truncated; two sections moved into the FAQ and everything shows again.
1.22.2
Fixed — Clicking a summary tile now lands the section in view. On the Discovery screen, jumping to Providers, Capabilities, APIs or Tools scrolled the section’s heading up under the pinned header; it now lands just below it, so you see the heading you asked for.
Fixed — No more “401” in the browser console. The MCP server settings card checked whether the server was answering by pinging its sign-in-only endpoint without credentials — a harmless 401 that the browser nonetheless logged as a red error on every admin page. It now asks over an authenticated route instead: same “Running” status, quiet console.
Improved — A tidier Discovery screen. The discovery documents were listed in three places at once; the endpoint card no longer repeats them. The always-visible rail keeps its quick links, and the “Well-known documents” section keeps the full inventory with each file’s generated/managed status — each with its own job.
Improved — AI Visibility explains its own keys. The settings now say, in place, why AI Visibility connects to each engine directly instead of using WordPress’s shared AI connector: a visibility check is graded on the sources each engine cited and compares several engines at once — neither of which the shared connector can provide.
1.22.1
New — Connected AI tools can write now — each step behind its own switch. The MCP server card grows a trust ladder. A second switch (off by default) gives connected agents five write tools: draft a new post or page, edit an existing one, set a page’s AI description and its Topics for AI, and apply a Readiness fix — where “fix” means exactly what the readiness check itself recommends, and only ever enabling a plugin feature; a fix that needs your judgement or a server change refuses honestly and names the manual step instead. A third switch decides publishing: while it’s off, an agent can only leave drafts and pending posts for your review. Every write runs as the WordPress user behind the key — an author’s key can’t do editor things, a new category still needs the category permission, exactly as in wp-admin — and every write is recorded under More → Agent access, attributed to the key that made it. With the writes switch off the tools aren’t just refused, they don’t exist on any surface — MCP, the abilities endpoint, discovery — so “read-only” stays literally true; and switching the MCP server off switches writes (and publishing) off with it, so the ladder can never disagree with itself.
New — An agent can dress a post completely. Creating and editing now also carry categories and tags (by name, with wp-admin’s exact permission rules: any author can create a tag, only category managers can create a category) and the featured image — an attachment you already have, or an image URL the plugin imports into your media library, alt text included, gated on the user’s upload permission and your site’s file-type and size rules. Draft it, dress it, leave it for review: no wp-admin visit required.
New — The MCP card connects your AI tool for you. Pick the tool you use — Claude Desktop, ChatGPT/Codex, Claude Code — and the card mints a named application password in one click (or takes one you saved), hands you a finished setup with the endpoint and login already woven in, and proves the connection with a built-in test.
Improved — The score card says how much, not just what’s next. Each rung now carries a quiet “n to fix” count, an unmeasured Cited rung says “not measured yet” instead of a dash that read as broken, and the “Next:” line explains why in a tooltip. Grading also stops nagging about pages nobody authored — a shop’s cart and checkout pages, and container pages whose only content is a plugin’s shortcode — so the Optimize rung measures your writing, not your plugins’.
Fixed — The Discovery hub explains its own numbers. Its tiles reconcile what your site has with what anonymous agents are shown — “3 providers — 1 public · 2 sign-in only” instead of a bare “1” above a three-row list. The “MCP & tools” card now states one total and splits it by provider — “Agentimus abilities: 14 · Core abilities: 2”, adding up in plain sight — with the doors that serve each group (the MCP server, the Abilities API route) named on the group itself and their endpoints listed without counts, so there are no two overlapping numbers left to wrongly add together. And the card no longer claims “no agent tools yet” when every tool is simply sign-in-only and deliberately kept out of the public documents. One genuine counting bug fell out of the recheck too: an ability hidden from REST — which no agent can run remotely — no longer inflates any tool count, so the numbers shown are the numbers the endpoints actually answer. And the Content types setting now previews, live, exactly which read capabilities your selection advertises — a ticked type brings its public taxonomies along (posts add categories and tags) — so the capabilities number finally traces back to the switch that produces it.
Fixed — WebMCP follows the browser’s move. Chrome relocated the WebMCP entry point (navigator.modelContext → document.modelContext); the search tool now registers on the current name and still recognises the old one, so it keeps working as browsers retire the alias.
1.22.0
New — An MCP server, one switch, nothing extra to install. Turn on Settings → Discovery → MCP server and AI tools you already use — Claude Code, Claude Desktop, Cursor, Codex — can talk to your site over the Model Context Protocol and run Agentimus’s nine read-only tools: readiness, AI traffic, request log, bot identification, and the page/JSON-LD/Markdown previews. The WordPress MCP Adapter library now ships inside the plugin and is loaded only when you opt in — with the switch off, nothing is loaded and no endpoint exists. Nothing becomes public: every call must sign in with a WordPress login (an application password works), every tool keeps the same permission checks as the admin screen it comes from, every call is recorded under More → Agent access — and switching it off disconnects connected tools immediately. To keep the surface exactly what you opted into, Agentimus also keeps the library’s generic “run any ability” server switched off. Needs WordPress 6.9+ (the Abilities API); off by default. Full transparency: the settings card also tells you what an application password really is — a login for your whole REST API, not just this server — and recommends one password per tool, on a user with only the permissions it needs.
Fixed — The MCP wiring had quietly gone stale — in two places. The adapter library renamed its transport class some versions ago; Agentimus still asked for the old name, so even a site that had installed the adapter by hand got no server at all, silently. And the mcp.json manifest read each server’s tool list with equally outdated accessors, so a running server was advertised as having “0 tools” and its per-server card refused to serve. Both fixed: the manifest now lists every server’s real tools and links a card for each.
New — Agent access now says who. Every row carries the user and the named application password behind it — “by anna · app password “zapier”” — resolved live from your own users and keys, so a renamed key shows its current name, a revoked one says “since revoked”, and a deleted user’s row says that too instead of a bare number. Password-lifecycle rows say “on anna’s account”, because that’s what is actually recorded — the key’s owner, not necessarily who clicked. Nothing new is stored: no IP addresses, no identities; the names are looked up at view time.
Changed — The dashboard’s 7- and 30-day numbers are whole calendar days now. They used to be rolling windows ending at the current second, which meant they could visibly shrink between midnights as week-old hits aged out — watched live under auto-refresh, that read as data loss. Every window is now counted in whole calendar days (UTC), the same clock as the Today tile and the daily chart beneath them; numbers move only when a hit arrives or at midnight UTC. The tiles may read slightly higher after updating — the window now includes its first day’s early hours, which rolling had already dropped.
