fimo 2.9.0 → 2.10.0-staging.10

package/assets/skills/fimo-studio/references/ai-chat.md

@@ -8,11 +8,11 @@ Chat has its own **Chats** tab in the site header; the tab shows how many chats
 
 ## Agents
 
-Agents and routines appear only when the organization has the agents feature, which is off by default in production. Without it, the **+** above the chat list opens a new chat directly, agent settings have no **Routines**, and existing routines do not run; ask Fimo support to enable it. The plan decides the agents (FIMO-2856): Free has the SEO and GEO default agents, each running only its weekly checkup (**Weekly SEO checkup**, **Weekly AI check**), and its agent settings list the routines Starter adds with **See plans**; Starter (pay-as-you-go) has every default agent with its detailed routines, and a change wakes each routine at most once a day (the changes made meanwhile wait and the next run, a day after the last one started, checks them together; the routine's next run shows when, and **Run now** and schedules never wait); Pro runs them right after each change and adds agents of one's own (**New agent**). In the **+** menu, what the plan does not include shows **Upgrade** and opens the plans. Default agents are Fimo's: their routines follow Fimo's instructions, only switch off and on or get deleted (a deleted one stays deleted), and nothing is added to them (`routine_is_default`); routines of one's own live in a custom agent, on Pro (`routine_needs_pro` below Pro). A default agent's name, description and portrait cannot be changed on any plan, and a custom agent's change only on Pro: below Pro its settings are read-only, say its routines are paused and offer **See plans**, and it can still be deleted. A plan change changes which agents open and which routines run; agents, routines and past runs stay. Fimo support can give an organization Pro's agents on any plan. A routine Fimo retires from a default agent stops running and leaves the list.
+The plan decides the agents (FIMO-2856): Free has the SEO and GEO default agents, each running only its weekly checkup (**Weekly SEO checkup**, **Weekly AI check**), and its agent settings list the routines Starter adds with **See plans**; Starter has every default agent with its detailed routines, and a change wakes each routine at most once a day (the changes made meanwhile wait and the next run, a day after the last one started, checks them together; the routine's next run shows when, and **Run now** and schedules never wait); Pro runs them right after each change and adds live agents of one's own (**New agent**: any plan can set one up, only Pro makes it live). In the **+** menu, what the plan does not include shows **Upgrade** and opens the plans. Default agents are Fimo's: their routines follow Fimo's instructions, only switch off and on or get deleted (a deleted one stays deleted), and nothing is added to them (`routine_is_default`); routines of one's own live in a custom agent, on Pro (`routine_needs_pro` below Pro). A default agent's name, summary, instructions and portrait cannot be changed on any plan, and a live custom agent's change only on Pro: below Pro its settings are read-only, say its routines are paused and offer **See plans**, and it can still be deleted. A plan change changes which agents open and which routines run; agents, routines and past runs stay. Fimo support can give an organization Pro's agents on any plan. A routine Fimo retires from a default agent stops running and leaves the list.
 
-An agent is a chat with a job: a name and a description that Fimo reads at the start of every reply in that chat. Agents are shared like any chat and are listed as soon as they are created, with an **Agent** tag after the name. The **+** above the chat list is a menu: **New chat** (blank), **New agent** (a name and a description, then the empty agent opens), and under **Agents** Fimo's default agents (**SEO**, **GEO**, **Content**, **Quality**, **Compliance**); a new site starts with SEO and GEO listed, their routines active from the start; Content, Quality and Compliance, and every default on an older site, are created the first time someone selects them, routines active, the same chat every time after. In an agent, the header shows the name alone and the empty state introduces the agent by name and job (the description's first three lines); a cut description shows whole on hover, or on tap on a phone, and the message box reads **Message <name>**. The agent's settings page has a **Description** field, shared with everyone; empty means none. The filter icon's **Kind** option shows only chats or only agents; search finds an agent by its name.
+An agent is a chat with a job: a name, a one-line summary for people, and instructions that Fimo reads at the start of every reply in that chat and every run of its routines (FIMO-3079). Live agents are shared like any chat, with an **Agent** tag after the name. The **+** above the chat list is a menu: **New chat** (blank), **New agent** (opens the new agent's chat at once: the agent asks what it should do with three one-click answers, then drafts its name, summary, instructions and routines on one card; **Create agent** sends the person's agreement and the agent goes live, the team sees it and its routines start; until then only its creator sees it, as **New agent** with **Setting up…**, its routines do not run, and its chat changes nothing on the site (it reads the site, its data and the web; no CLI, files, work sessions or publish), and it goes live only after the person saw its latest draft; a routine time the person did not name keeps its hour and Fimo picks the minute; below Pro, **Create agent** opens the plans), and under **Agents** Fimo's default agents (**SEO**, **GEO**, **Content**, **Quality**, **Compliance**); a new site starts with SEO and GEO listed, their routines active from the start; Content, Quality and Compliance, and every default on an older site, are created the first time someone selects them, routines active, the same chat every time after. In an agent, the header shows the name alone and the message box reads **Message <name>**. A default agent's empty state greets in its own words and offers its own starter requests; its description stays in its settings. A starter is sent as shown, by anyone who can send messages. Asked to, an agent changes its own instructions with its `agent` tool; the chat shows **Updated instructions** with **View change** (before and after). The agent's settings page has **Summary** and **Instructions** fields, shared with everyone (empty means none), and an **Instruction history** where **Restore previous** puts back the text a change replaced. In text, an agent links another as `[Name](agent://id)`, shown as a chip that opens it. The filter icon's **Kind** option shows only chats or only agents; search finds an agent by its name.
 
-The **SEO** default runs six routines backed by Studio's built-in `fimo-seo` skill, which checks the site as a browser renders it and fixes at the source: **Technical SEO** (status, redirects, canonicals, sitemap, robots.txt, language, main heading, and changes since its last run: pages moved without a redirect), **Broken links**, **Meta tags** (titles, descriptions, Open Graph, the site's title ending, length in characters and pixels), **Image SEO** (alt text, file names, images over 200 KB: one in the media library or the site's own files is rendered through Fimo's resizing on its own, a file of the site uploaded first), **Hreflang** (only on a site in more than one language) and **Internal linking** (orphan and deep pages, articles with no link up to a hub or product page or more than eight links). Each default routine wakes on the changes on main that can matter to it, before a publish (CMS edits, code changes, images, languages, forms; the first publish and a new domain for Technical SEO and Cookie consent): a small model reads each change first, so a quiet site runs nothing, and the run checks what changed (the whole site on its first run). Only what changes without anyone touching the site keeps a schedule: Broken links checks outbound links on the 1st of every month, Content refresh runs on the 2nd, Content lifecycle every three months, Privacy and legal pages once a year. A page it changed is left alone for 8 weeks, the time Google takes to show the effect. A fix with one right answer from the page's own words is made without a message (a link only once the new target is checked to be the same content in the same language; a link on words the text already uses to name another page; a 301 from a page's old address once the same page is live at a new one); a change to what visitors read, which page wins a search, whether a page exists, a redirect with more than one possible target, a price or a legal page is asked as a decision; page removals and the robots policy are never changed on its own. A change a person asks for in a chat is made. What a check could not tell (a timeout, a 5xx, a page past the crawl limit) is never acted on: the run retries once, then says the check was incomplete.
+The **SEO** default runs six routines backed by Studio's built-in `fimo-seo` skill, which checks the site as a browser renders it and fixes at the source: **Technical SEO** (status, redirects, canonicals, sitemap, robots.txt, language, main heading, changes since its last run: pages moved without a redirect, and, with Search Console, old addresses Google still shows that the site no longer serves), **Broken links**, **Meta tags** (titles, descriptions, Open Graph, the site's title ending, length in characters and pixels), **Image SEO** (alt text, file names, images over 200 KB: one in the media library or the site's own files is rendered through Fimo's resizing on its own, a file of the site uploaded first), **Hreflang** (only on a site in more than one language) and **Internal linking** (orphan and deep pages, articles with no link up to a hub or product page or more than eight links). Each default routine wakes on the changes on main that can matter to it, before a publish (CMS edits, code changes, images, languages, forms; the first publish and a new domain for Technical SEO and Cookie consent): a small model reads each change first, so a quiet site runs nothing, and the run checks what changed (the whole site on its first run). A schedule covers what no single change shows: every Monday, Technical SEO, Meta tags and Internal linking read every page of main's hosted preview (pages out of the sitemap, shared titles and title shape, orphan and deep pages); what changes without anyone touching the site also keeps one: Broken links checks outbound links on the 1st of every month, Content refresh runs on the 2nd, Content lifecycle every three months, Privacy and legal pages once a year. A page it changed is left alone for 8 weeks, the time Google takes to show the effect. A fix with one right answer from the page's own words is made without a message (a link only once the new target is checked to be the same content in the same language; a link on words the text already uses to name another page; a 301 from a page's old address once the same page is live at a new one); a change to what visitors read, which page wins a search, whether a page exists, a redirect with more than one possible target, a price or a legal page is asked as a decision; page removals and the robots policy are never changed on its own. A change a person asks for in a chat is made. What a check could not tell (a timeout, a 5xx, a page past the crawl limit) is never acted on: the run retries once, then says the check was incomplete.
 
 The **GEO** default runs three routines backed by Studio's built-in `fimo-geo` skill, which shares `fimo-seo`'s crawl and its rule for what a run may change alone: **AI readiness** (llms.txt, the text each page serves before JavaScript, landmarks and heading levels) and **Structured data** (valid JSON-LD; Organization and WebSite, Article, BreadcrumbList; Product and FAQ where they fit) and **AI citation optimization** (a direct answer under each content page's H1, question headings on long pages). On its own it writes a missing llms.txt and keeps its links in step, adds Organization, WebSite, Article and BreadcrumbList from fields the site already has (never an invented logo, address, author or rating), and fixes landmarks and heading levels without changing words; an opening answer, new headings, FAQ or Product data, and how pages are rendered are asked as decisions; who robots.txt lets crawl, AI crawlers included, prices and claims are never changed, nor reported: asked whether an assistant can read the site, it checks and names the crawlers robots.txt keeps out. Where AI assistants cite the site needs a visibility source; until one is connected there is no routine for it.
 
@@ -24,7 +24,7 @@ The **Compliance** default runs three routines backed by Studio's built-in `fimo
 
 ### Routines
 
-A routine is saved instructions an agent runs in a thread of its own, by hand, on a schedule, or when something happens in the project, so each run sees its earlier runs and the agent's chat never fills with its work. The agent creates, changes, runs and deletes its routines from its chat with its `routines` tool: a person asks in the chat ("check the prices every Monday", "run after each publish"), and Studio only shows routines. A trigger is set in words; Fimo checks them and either says when the routine runs or why it cannot follow them and what it can do instead; for what Fimo cannot see as an event (traffic, rankings, errors) the agent offers a schedule whose run checks it and stops when there is nothing to do. Fimo follows publishes of the site (or only its first publish) and content changes on the main branch by content type, action, locale, changed fields, the entry's current values ("price over 100", "status changes to archived") or a yes/no question about what the content says ("mentions Contentful or Sanity"; for "a competitor" it uses the names given or those in the site knowledge, and asks for them otherwise; a run started on an uncertain answer says so), media files uploaded or changed (by name or file type), the site's code changing through a deploy or a merge (by file path glob, or a yes/no question about the changed files such as "a page was added"), another routine of the same agent finishing (optionally with a yes/no question about its report; never its own end), a visitor submitting a form (optionally with a yes/no question about the answers), a form or content type changing, a branch merged into main, a locale added, removed or made the default, a publish of the site failing, a domain added or removed, and the prices on a page of another web site changing ("tell me when https://example.com/pricing changes its prices": checked daily, a change must hold on three reads before the routine runs, the run lists each plan with its old and new price, the routine page shows the watch as **OK**, **Needs repair**, **Blocked** or **Unstable** with its last check, at most 10 watched pages a project); comparing a field's old and new values is not supported; what a routine changes, publishes or reports can start the agent's other routines, in chains of any length, each routine at most once per change that started the chain (across every routine that change leads to, so routines never loop or multiply; a person's new change, a run by hand or on a schedule starts a new chain); events that come in while a run is busy wait, then run together in one next run (up to 10), which the Runs list shows as one run with its event count; the agent can also make a routine's event runs wait until the changes stop for a set number of minutes (at most 2 hours from the first), then read them all in one run (a yes/no question is then answered once for all the waiting changes, each entry once in its latest version; a waiting change it says no to leaves the queue, and when it says no to all of them no run starts); the default routines that read what changed (Meta tags, Image SEO, Internal linking, Structured data, AI citation optimization, Content lifecycle, Translation, Accessibility, Mobile optimization, Brand compliance) wait 10 minutes after the last change, and those that guard the live site (Technical SEO, Broken links, Hreflang, AI readiness, Cookie consent, Privacy and legal pages) run at once. The agent's settings list its **Routines** as cards (name, when it runs such as its event trigger's title, **Every Monday at 08:00** or **By hand**, next run, an **Active** switch); the default agents come with theirs (a new project's defaults wake on changes from five minutes after it is created, so its first push and seed start no run). A routine's page is read-only: **Instructions** (the task, run on every trigger), **Model** (the model its runs use: Opus 5.5 by default, while most default routines run on Claude Haiku 5.5 or Claude Sonnet 5; the agent sets another when asked), **Triggers** (**By hand** with **Run now**, which starts a run right away; **Schedule** and the event trigger, given a short title by Fimo from the words it was set with, each a sentence, shown when set), the **Runs** list (newest first, two lines each: how it ended (**Quiet**, **Sent to agent**, **Asked for a decision**, or **Failed** and **Stopped** in red), when and what started it (**On schedule**, **On event**, or **By hand** with the person's name or **agent** when their message started it), then the first sentence of what it told the agent; select one to open the routine's thread, read-only, at that run), and the **Active** switch in the header (pauses the schedule and the trigger, not **Run now**). Routines report to the agent, never to a person: a run that has something the agent should know writes it to the agent's chat as a message only the agent reads, and the agent takes a turn on it (reports waiting together are read in one turn); a run with nothing to say reports nothing. The agent decides what the team sees: nothing (the chat stays as it was, the row stays read), one short message (several reports folded into one), or a form when a person must decide, and the chat's row reads **Needs you** while that form is open. A person answers the form in the chat; the agent sends the decision to the routine with its `routines` tool and the routine runs again with it. Only the agent answers a routine: its page has no answer box. Quiet runs leave no message: they fold into one line placed where the last run happened among the messages (**N runs since <date>**), counted from the last report the agent answered visibly, and the chat's row reads **Ran <date>** with that count while the newest thing in the chat is a run, the ordinary status and the last words when a message is newer, and **First check <day> <time>** or **No checks scheduled** before the first run. A routine whose runs keep failing or are refused for credits is reported to the agent once per streak, as a message from Fimo. Routine runs use the plan's agent-hours, from each run's start to its end (Studio chats do not); Billing shows the hours used this period, and each run shows its duration and credits. Once the included hours are used, nothing runs (no schedule, event or Run now, which the routines tool reports to the agent); Studio shows **Routines paused** under the chat list and in the agent's settings with the reset date and the button that resumes them sooner (**Upgrade**), and the routine's page says until when. No run lasts longer than 2 hours.
+A routine is saved instructions an agent runs in a thread of its own, by hand, on a schedule, or when something happens in the project, so each run sees its earlier runs and the agent's chat never fills with its work. The agent creates, changes, runs and deletes its routines from its chat with its `routines` tool: a person asks in the chat ("check the prices every Monday", "run after each publish"), and Studio only shows routines. A trigger is set in words; Fimo checks them and either says when the routine runs or why it cannot follow them and what it can do instead; for what Fimo cannot see as an event (traffic, rankings, errors) the agent offers a schedule whose run checks it and stops when there is nothing to do. Fimo follows publishes of the site (or only its first publish) and content changes on the main branch by content type, action, locale, changed fields, the entry's current values ("price over 100", "status changes to archived") or a yes/no question about what the content says ("mentions Contentful or Sanity"; for "a competitor" it uses the names given or those in the site knowledge, and asks for them otherwise; a run started on an uncertain answer says so), media files uploaded or changed (by name or file type), the site's code changing through a deploy or a merge (by file path glob, or a yes/no question about the changed files such as "a page was added"), another routine of the same agent finishing (optionally with a yes/no question about its report; never its own end), a visitor submitting a form (optionally with a yes/no question about the answers), a form or content type changing, a branch merged into main, a locale added, removed or made the default, a publish of the site failing, a domain added or removed, and the prices on a page of another web site changing ("tell me when https://example.com/pricing changes its prices": checked daily, a change must hold on three reads before the routine runs, the run lists each plan with its old and new price, the routine page shows the watch as **OK**, **Needs repair**, **Blocked** or **Unstable** with its last check, at most 10 watched pages a project); comparing a field's old and new values is not supported; what a routine changes, publishes or reports can start the agent's other routines, in chains of any length, each routine at most once per change that started the chain (across every routine that change leads to, so routines never loop or multiply; a person's new change, a run by hand or on a schedule starts a new chain); events that come in while a run is busy wait, then run together in one next run, which the Runs list shows as one run with its event count (a very large burst, such as a bulk import, runs in several runs, in order); the agent can also make a routine's event runs wait until the changes stop for a set number of minutes (at most 2 hours from the first), then read them all in one run (a yes/no question is then answered once for all the waiting changes, each entry once in its latest version; a waiting change it says no to leaves the queue, and when it says no to all of them no run starts); the default routines that read what changed (Technical SEO, Meta tags, Image SEO, Internal linking, AI readiness, Structured data, AI citation optimization, Content lifecycle, Translation, Accessibility, Mobile optimization, Brand compliance) wait 10 minutes after the last change, and those that guard the live site (Broken links, Hreflang, Cookie consent, Privacy and legal pages) run at once. The agent's settings list its **Routines** as cards (name, when it runs such as its event trigger's title, **Every Monday at 08:00** or **By hand**, next run, an **Active** switch); the default agents come with theirs (a new project's defaults wake on changes from five minutes after it is created, so its first push and seed start no run). A routine's page is read-only: **What it does** (one or two plain sentences the agent writes with the routine, Fimo's for a default routine), with **Instructions** folded under it (the task, run on every trigger; a routine created before descriptions shows only its **Instructions**), **Model** (the model its runs use: Opus 5.5 by default, while most default routines run on Claude Haiku 5.5 or Claude Sonnet 5.5; the agent sets another when asked), **Triggers** (**By hand** with **Run now**, which starts a run right away; **Schedule** and the event trigger, given a short title by Fimo from the words it was set with and showing those words, shown when set), the **Runs** list (newest first, two lines each: how it ended (**Quiet**, **Sent to agent**, **Asked for a decision**, or **Failed** and **Stopped** in red), when and what started it (**On schedule**, **On event**, or **By hand** with the person's name or **agent** when their message started it), then the first sentence of what it told the agent; select one to open the routine's thread, read-only, at that run), and the **Active** switch in the header (pauses the schedule and the trigger, not **Run now**). Routines report to the agent, never to a person: a run that has something the agent should know writes it to the agent's chat as a message only the agent reads, and the agent takes a turn on it (reports waiting together are read in one turn); a run with nothing to say reports nothing. The agent decides what the team sees: nothing (the chat stays as it was, the row stays read), one short message (several reports folded into one), or a form when a person must decide, and the chat's row reads **Needs you** while that form is open. A person answers the form in the chat; the agent sends the decision to the routine with its `routines` tool and the routine runs again with it. Only the agent answers a routine: its page has no answer box. Quiet runs leave no message: they fold into one line placed where the last run happened among the messages (**N runs since <date>**), counted from the last report the agent answered visibly, and the chat's row reads **Ran <date>** with that count while the newest thing in the chat is a run, the ordinary status and the last words when a message is newer, and **First check <day> <time>** or **No checks scheduled** before the first run. A routine whose runs keep failing or are refused for credits is reported to the agent once per streak, as a message from Fimo. Routine runs use the plan's agent-hours, from each run's start to its end (Studio chats do not); Billing shows the hours used this period, and each run shows its duration and credits. Once the included hours are used, nothing runs (no schedule, event or Run now, which the routines tool reports to the agent); Studio shows **Routines paused** under the chat list and in the agent's settings with the reset date and the button that resumes them sooner (**Upgrade**), and the routine's page says until when. No run lasts longer than 2 hours.
 
 ## Sending a Message
 
@@ -37,7 +37,7 @@ A routine is saved instructions an agent runs in a thread of its own, by hand, o
 
 The AI can place choices, cards, tables, images, and contextual actions directly in its reply. Optional suggestions do not interrupt the conversation.
 
-When the AI needs an answer before it can continue, it pins a form directly above the message box and waits for your response. While the question is the last message, the form stays there as you scroll the conversation; the message that asked shows **Waiting for your answer...** in its place in the history. If you send a message that does not answer it, the question moves back to its place in the history, folded to one line with an orange **?**. Select it to open it there and answer it at any time, with an optional note from **Add a note**; the answer is sent as a new message at the end of the chat. An answered question shows a gray check next to its title. Select **Minimize** to fold the form to one line without losing your answers, then select that line to reopen it. **Submit** sends only the completed form and leaves anything typed in the message box unsent. Sending from the message box includes any selections already made in the form with your message; with no selections, it is an ordinary reply and the question stays open. A form with many steps scrolls inside its own area, and its **Next** / **Submit** buttons stay in view.
+When the AI needs an answer before it can continue, it opens a form in the conversation, under the message that asks, and waits for your response. The form goes one step at a time with **Previous**, **Next**, and **Submit**, and a step with choices has an optional note field under them. If you send a message that does not answer it, the question folds to one line with an orange **?** where it was asked. Select it to open it there and answer it at any time; the answer is sent as a new message at the end of the chat. An answered question shows a gray check next to its title. When a routine asks again before anyone answered, its earlier question stays unanswered, folded where it was asked. Select **Minimize** to fold the form to one line without losing your answers, then select that line to reopen it. **Submit** sends only the completed form and leaves anything typed in the message box unsent. Sending from the message box includes any selections already made in the form with your message; with no selections, it is an ordinary reply and the question stays open.
 
 ## Attaching Context
 

package/assets/skills/fimo-studio/references/billing-and-credits.md

@@ -34,7 +34,7 @@ Buying credits is available while a paid plan is effective, including after sche
 1. Click the credit display near the chat input.
 2. Click **Buy credits**.
 3. Enter a dollar amount (up to $10,000) or pick one of the presets: **$25**, **$50**, **$100**, or **$200**.
-4. Check the rate shown in the dialog. It comes from your current plan, so the same dollar amount can buy a different number of credits on Pay as you go than on Pro.
+4. Check the rate shown in the dialog. It comes from your current plan, so the same dollar amount can buy a different number of credits on different plans.
 5. Confirm the purchase. Your card on file is charged and the credits are added after confirmation; there is no separate checkout page to complete. If confirmation stays pending, retry the same purchase or contact support. Do not start another purchase to replace an unresolved one.
 
 ## Refilling Credits Automatically
@@ -57,29 +57,38 @@ You can upgrade from two places:
 From there:
 
 1. Compare the available plans side by side.
-2. Pro includes more monthly credits, custom domain support, and the ability to hide the Fimo badge on your published site.
+2. Starter and Pro include more monthly credits and custom domain support. Pro has the largest allowances and the ability to hide the Fimo badge on your published site.
 3. Pick a plan and complete checkout.
 4. Your new monthly credit pool is available right away.
 
+## Activating a team organization
+
+Only your personal organization starts on Free. A new team organization opens nothing until someone completes a Starter or Pro checkout: every page of it lands on **Choose a plan for {organization}**, and its sites, integrations, members, domains and settings stay unavailable until then.
+
+- An owner or admin picks **Continue with Starter** or **Continue with Pro** and completes checkout. Other members see that an owner or admin must activate it.
+- Leaving checkout without paying keeps the organization inactive. Use the organization switcher at the top of the page to go back to another organization.
+- An owner who no longer wants the organization can delete it from the same page.
+- A team organization that has paid once stays usable if it later moves to Free.
+
 ## Extra usage budget
 
-Open **Billing & usage**, then **Details** beside public bandwidth. Billing admins on Pay as you go or Pro can edit one **Monthly budget** for extra public bandwidth. The page shows its cost and **Spent this period**. This budget does not cover AI credit purchases or auto-refill.
+Open **Billing & usage**, then **Details** beside public bandwidth. Billing admins on Starter or Pro can edit one **Monthly budget** for extra public bandwidth. The page shows its cost and **Spent this period**. This budget does not cover AI credit purchases or auto-refill.
 
 - At the budget, additional paid usage stops; remaining included allowances stay usable.
 - Disabling extra usage preserves charges already incurred. The budget cannot be lowered below existing extra-usage charges.
 
 ## Changing or leaving a paid plan
 
-- Free to Pay as you go or Pro, and Pay as you go to Pro, use checkout and take effect after payment succeeds.
-- Moving down from a paid monthly plan stops auto-refill and paid extra usage immediately, whether the destination is Free or Pay as you go. The current plan, remaining allowances, credits, and reset date stay unchanged until the destination takes over at renewal. Manual credit purchases remain available while the paid plan is active.
+- Free to Starter or Pro, and Starter to Pro, use checkout and take effect after payment succeeds. Starter to Pro credits back the unused part of the Starter month.
+- Moving down from a paid monthly plan stops auto-refill and paid extra usage immediately, whether the destination is Free or Starter. The current plan, remaining allowances, credits, and reset date stay unchanged until the destination takes over at renewal. Manual credit purchases remain available while the paid plan is active.
 - Automatic spending cannot be enabled while a plan change is scheduled. Keep your plan first; then you can immediately enable auto-refill and paid extra usage independently. Keeping the plan itself leaves both off.
 - Usage before the change remains payable. Enabling extra usage again does not make usage from the off-period billable.
-- Pro to Pay as you go or Free takes effect at the end of the current billing period. Billing shows the destination and effective date until then.
-- Review the confirmation before scheduling Pro to Pay as you go. It explains the immediate spending stop, retained benefits and charges already incurred; opening it does not change the plan.
-- Switching the scheduled destination between Free and Pay as you go keeps the same effective date and the same moment automatic spending stopped. Nothing is re-granted. Team organizations reach Free through **Cancel subscription**, not the plan dialog.
-- Choosing the current plan again before that date cancels the scheduled change and its Pay as you go successor.
-- Pay as you go requires an active payment method. Removing its only valid card returns the organization to Free immediately, and a scheduled Pay as you go with no card on file lands on Free on its effective date instead.
-- A failed Pro renewal follows the payment-recovery flow. It never silently changes the organization to Pay as you go.
+- Pro to Starter, or Pro or Starter to Free, takes effect at the end of the current billing period. Billing shows the destination and effective date until then.
+- Review the confirmation before scheduling Pro to Starter. It explains the immediate spending stop, retained benefits and charges already incurred; opening it does not change the plan.
+- Switching the scheduled destination between Free and Starter keeps the same effective date and the same moment automatic spending stopped. Nothing is re-granted. Team organizations reach Free through **Cancel subscription**, not the plan dialog.
+- Choosing the current plan again before that date cancels the scheduled change and its Starter successor.
+- A scheduled Starter starts with its first monthly payment. With no card on file that payment fails and the organization moves to Free through payment recovery.
+- A failed Pro renewal follows the payment-recovery flow. It never silently changes the organization to Starter.
 
 ## Legacy plan transitions
 

package/assets/skills/fimo-studio/references/integrations.md

@@ -0,0 +1,105 @@
+# Integrations
+
+Integrations let a site use external services without exposing credentials to editors, chat, or agents. Studio, the CLI, and
+agent chats all resolve the same integration state.
+
+## Connect an integration for the organization
+
+Open **Integrations** in the organization sidebar. It lists the integrations the organization already connected, then the ones still
+available, with the number of sites each integration is on. Choose an integration to open its panel.
+
+An organization admin connects an integration once for the whole organization, with no site needed:
+
+- **Google Search Console** and **Google Analytics**: **Sign in with Google**, using a Google account that can see the
+  site's property, and approve the permissions Fimo lists.
+- Integrations you sign in to: Ahrefs, DataForSEO, PostHog, Resend, Calendly and Cal.com. **Connect <integration>**
+  opens the provider's sign-in page and returns to Fimo. Finish in the same browser, signed in to Fimo as yourself.
+  Ahrefs, PostHog and Resend also offer **Use an API key instead**.
+- Semrush connects with a Semrush API key: **Connect Semrush** opens the form with the `Authorization` header and
+  `Apikey ` prefix filled in.
+- Slack: **Add to Slack** (in the organization's Integrations or a site's Integrations; each sign-in adds a workspace), then choose each
+  site's channels.
+
+People who are not organization admins see **Copy link** instead, to send its link to an admin. The panel also shows
+the connected account, with **Disconnect** (removes the integration from every site, with its settings and always-allowed tools), or **Reconnect** when the
+provider stopped accepting it. A custom MCP server's **…** menu has **Rename** and **Delete server**. An MCP integration's
+panel also shows its server and tools: see [Manage an MCP server](#manage-an-mcp-server).
+
+### Site access
+
+Under **Site access** in the integration's panel, choose **All sites** to give every site access, including sites created later,
+or **Only select sites** and **Select sites**. Choose **Save**. A site left out loses the integration, and its managers cannot
+turn it back on until an admin gives the site access again.
+
+## Use an integration on a site
+
+Open the site, then **Settings** → **Integrations**. The page lists the integrations **On for this site** first, then the **Available**
+ones. Select an integration to open its panel. It says whether the integration is on for the site; the account, credentials and tools
+live in the organization's **Integrations**:
+
+- The switch turns the integration on or off for this site. People who manage the site can do this for any integration the site has
+  access to; others see it without changing it. An integration the organization has not connected says so and links to **Integrations**.
+- Settings appear once the integration is on. Google integrations ask for the **Search Console property** or **Google Analytics
+  property** the site uses; Fimo suggests the one that matches the site's domain. An integration still missing a setting shows
+  **Finish setup** on its card, and its panel asks for the setting first; connecting from the site opens it there.
+- **Manage the account and site access in Integrations** opens the organization page, for admins.
+
+## Approve a change
+
+Reads from an integration run straight away, and so do changes a person runs. When an agent wants to change data in an integration,
+the chat shows a card with the integration, the tool and why it asks:
+
+- **Allow** runs that exact change once.
+- **Always allow** (for changes that are not paid actions) runs it and lets agents run that tool on this site
+  without asking again, in any chat.
+- **Decline** stops it.
+
+Paid actions ask every time. A routine's request appears as a permission card in its agent's chat, with
+**Always allow on this site** and **Decline**. Allowing it starts another run of that routine with the original action;
+declining it does not. Do not make a separate question or ask the person to visit Settings for this approval.
+Changes asked from Slack, the CLI or the hosted MCP give a link to **Settings** → **Integrations**, where the same card appears.
+For Slack, Fimo continues in the thread once it is allowed. For
+the CLI or an AI client, **Allow** shows a one-time code to paste back; after **Always allow** the same call runs without
+one. A request is answered once and an approval runs one exact change, both within an hour.
+
+## Slack channels
+
+Open **Slack**, then **Workspaces and channels**. Each Slack workspace the site answers in has a row. Open one to turn
+it on for the site and choose the channels where Fimo answers mentions. Which sites use a workspace is Site access, and
+disconnecting it is the account's **…** menu, both on Slack's page in the organization's Integrations, where **Add another
+account** adds another workspace.
+
+## Add an MCP server
+
+Choose **Add MCP server** in a site's **Settings** → **Integrations** (anyone who manages the site), or on the organization's
+**Integrations** page (organization owners and admins). A server added from a site shows in the organization's Integrations too:
+
+1. Enter a **Name** and the server's **MCP URL**, a Streamable HTTP `https://` endpoint. A listed MCP integration, such as Semrush
+   or PostHog, fills both when you connect it from its panel.
+2. **Connect with** shows the recommended sign-in path. Fimo checks the endpoint; public servers connect immediately,
+   while OAuth servers open their provider sign-in page and return after approval.
+3. Open **Other ways to connect** when the server instead documents an access token, custom headers, or no
+   authentication.
+4. In a site, **Use from another site** reuses a server another site already connected.
+
+For **Paste an access token**, enter only the token; Fimo adds the `Bearer` scheme. For **Send custom headers**, enter
+each complete value exactly as the server expects, including any prefix (`Authorization` with `Bearer abc123`). Fimo
+encrypts every saved grant and never shows its private value again.
+
+## Manage an MCP server
+
+An MCP integration's panel in the organization's **Integrations** shows the server's address and its tools, grouped into reads,
+changes and deletes; open a group, then a tool, for its full description. When the tools stop loading, the panel offers
+the one repair for how the organization signs in:
+
+- **Reconnect** signs in again for OAuth.
+- **Update credentials** replaces a bearer token or the complete custom-header bundle after it stops working.
+- **Retry connection** asks a public server for its tools again without touching any credential.
+
+The account's **…** menu also has **Rotate credentials**, to replace a healthy token or header bundle. A custom
+server's **…** menu has **Delete server**, which removes the server from the organization, for every site that uses it.
+Only organization owners and admins can replace a shared credential or delete a server.
+
+Only integrations turned on for the site are available to Studio chat and the Fimo CLI for this site. Agents should discover
+them with `fimo integrations list`, inspect one with `fimo integrations info <integration>`, and run an allowed tool with
+`fimo integrations run <integration> <tool>`.

package/assets/skills/fimo-studio/references/project-creation.md

@@ -11,13 +11,13 @@ To build with your own AI instead, select **Build with ChatGPT, Claude or a codi
 ## Start From a Website
 
 1. Enter the website address. Fimo checks that it can read the site; if it cannot, it says why and nothing is created.
-2. Fimo creates the site, reads its homepage, works out what kind of site it is (business, non-profit, portfolio, hobby or personal site), and asks three questions that fit it in a short chat. It always starts by confirming what it understood. A business then keeps or adds its competitors and says how it stands out and how it is priced; other sites say what visitors should do and what to put forward, picking suggestions or typing their own. The answers are saved as the site's knowledge. If Fimo cannot read the site for the questions, select **Try again**.
-3. The report opens as soon as the last question is answered. It shows how ready the site is for search engines (SEO) and AI answers (GEO), with every check and how to fix it. Leaving and coming back resumes where you left off.
-4. Select **Copy & optimize my homepage**, at the top of the report. The site opens on its **Getting started** chat, where Fimo rebuilds the homepage with the report's problems fixed, then shows the scores before the rebuild and the checks it fixed. Only the homepage is rebuilt; links to other pages still go to the current site. Fimo covers the cost of the analysis, the questions and the first two rebuilds; credits are not used for them.
+2. Fimo creates the site, reads its homepage, works out what kind of site it is (business, non-profit, portfolio, hobby or personal site), and asks three questions that fit it in a short chat. It always starts by confirming what it understood. A business then keeps or adds its competitors and says how it stands out and how its prices compare to those competitors; other sites say what visitors should do and what to put forward, picking suggestions or typing their own. The answers are saved as the site's knowledge. If Fimo cannot read the site for the questions, select **Try again**.
+3. The report opens as soon as the last question is answered. It shows how ready the site is for search engines (SEO) and AI answers (GEO), with every check and how to fix it. To keep or share it, select **Export**. Leaving and coming back resumes where you left off.
+4. Select **Recreate and fix my homepage**, under the score. The site opens on its **Getting started** chat, where Fimo rebuilds the homepage with the report's problems fixed, then shows the new score and what it changed. Only the homepage is rebuilt; links to other pages still go to the current site. The first two homepage rebuilds are free; after that, a rebuild uses credits.
 
 Fimo does not rebuild a site that breaks its terms of use: the onboarding stops with a message to contact support.
 
-The report and the homepage copy Fimo read are deleted after 7 days.
+The report and the homepage copy Fimo read are deleted after 7 days. Export the report to keep it.
 
 Once the homepage is rebuilt, a banner offers to bring over the other pages: copy a prompt into an AI, or book a call or write to the Fimo team. Website migration is in alpha, so review every result. Dismissing the banner hides it for good.
 

package/dist/cli/bundle.json

@@ -1,7 +1,7 @@
 {
   "bundled": true,
   "bundler": "esbuild",
-  "bundledAt": "2026-10-21T15:49:00.346Z",
+  "bundledAt": "2026-10-28T08:55:06.555Z",
   "cliVersion": "2.9.0",
   "external": [
     "oxc-parser",