Pick any line to see why it is the way it is: the commit, the pull request and issue it came from, and what the agent was thinking.
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 1 | /** |
| 2 | * Home's parts (routes/workspace/home.tsx): Where you're needed, Start | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 3 | * here, what happened in the span (what people did, what landed, builds, |
| 4 | * decisions, spend) and Running now. Each takes what | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 5 | * lib/home.ts worked out and only draws it. |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 6 | */ |
| 7 | import { | |
| 8 | AlertTriangle, | |
| 9 | ArrowRight, | |
| 10 | AtSign, | |
| 11 | Bell, | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 12 | GitCommitHorizontal, |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 13 | GitMerge, |
| 14 | GitPullRequest, | |
| 15 | Lightbulb, | |
| 16 | Loader, | |
| 17 | Mail, | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 18 | Package, |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 19 | Rocket, |
| 20 | Server, | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 21 | Tag, |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 22 | User, |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 23 | Users, |
| 24 | Workflow, | |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 25 | } from "lucide-react"; |
| 26 | import type { ReactNode } from "react"; | |
| 27 | import { Link } from "react-router"; | |
| 28 | ||
| 29 | import { | |
| 30 | type Attention, | |
| 31 | type AttentionKind, | |
| 32 | type AttentionRow, | |
| 33 | type DecisionRow, | |
| 34 | type DeployRow, | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 35 | type Digest, |
| 36 | type DigestActor, | |
| 37 | type DigestColumn, | |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 38 | type LandedChange, |
| 39 | type RunningRow, | |
| 40 | type Span, | |
| 41 | type Spend, | |
| 42 | type Stake, | |
| 43 | WINDOWS, | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 44 | listOf, |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 45 | waited, |
| 46 | whenShort, | |
| 47 | } from "../lib/home"; | |
| 48 | import { cn } from "../lib/cn"; | |
| Money is written one way. A single formatter turns millionths of a dollar into dollars, rounding half up on whole micros rather than on a float, so the same sum reads the same on every page: under a cent reads <$0.01 on a total and exactly nothing is $0.00, while the statement's lines, a session's receipt, the price book and an agent's effort costs carry up to four places where the fraction of a cent is the point; the six formatters that each rounded their own way are gone. This month is the calendar month in UTC from its first day to today everywhere, and billing counts the month's not-yet-closed usage in any range that reaches into the current month, so the top bar's pill, Spend, Home and Usage ask for the same days and get the same figure; Home now reads the usage report the others read instead of adding up statements. Usage's pending sentence says what of that usage the close will charge after the discount and included usage, which the billing API returns as pending_charged_micros. A reconciliation test holds the pill, Spend, Home and Usage to one number for one month. The usage and billing guide says how amounts are written and what this month means. | 49 | import { money } from "../lib/money"; |
| The site is built from its shared components: one Button with its variants and sizes, one Badge, Card for every panel, Alert for notices, Kbd for keys, Table for the security tables, and Avatar, Command, Progress, ScrollArea, ToggleGroup, Collapsible and Accordion in the kit; raw buttons, hand-drawn cards and pills and per-file button class lists are gone, outline buttons stand the same height as filled ones, and the phone's Ask g1t loses its tile. Every page was compared pixel by pixel before and after. | 50 | import { ButtonLink } from "./ui"; |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 51 | import { Avatar } from "./ui/avatar"; |
| 52 | import { Badge } from "./ui/badge"; | |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 53 | import { Skeleton } from "./ui/skeleton"; |
| 54 | import { Tooltip, TooltipContent, TooltipTrigger } from "./ui/tooltip"; | |
| 55 | ||
| 56 | export function Card({ | |
| 57 | title, | |
| 58 | hint, | |
| 59 | link, | |
| 60 | children, | |
| 61 | className = "", | |
| 62 | }: { | |
| 63 | title: ReactNode; | |
| 64 | /** A few words under the title. */ | |
| 65 | hint?: string; | |
| 66 | link?: { label: string; to: string } | null; | |
| 67 | children: ReactNode; | |
| 68 | className?: string; | |
| 69 | }) { | |
| 70 | return ( | |
| 71 | <section className={`rounded-2xl border border-line bg-surface ${className}`}> | |
| 72 | <header className="flex items-start gap-2 px-5 pt-4 pb-3"> | |
| 73 | <div className="min-w-0"> | |
| 74 | <h2 className="text-sm font-semibold text-fg">{title}</h2> | |
| 75 | {hint && <p className="mt-0.5 text-xs text-muted">{hint}</p>} | |
| 76 | </div> | |
| 77 | {link && ( | |
| 78 | <Link to={link.to} className="ml-auto inline-flex shrink-0 items-center gap-1 pt-0.5 text-xs text-muted hover:text-fg"> | |
| 79 | {link.label} | |
| 80 | <ArrowRight size={12} /> | |
| 81 | </Link> | |
| 82 | )} | |
| 83 | </header> | |
| 84 | {children} | |
| 85 | </section> | |
| 86 | ); | |
| 87 | } | |
| 88 | ||
| 89 | export function Quiet({ children }: { children: ReactNode }) { | |
| 90 | return <p className="px-5 pb-5 text-sm text-muted">{children}</p>; | |
| 91 | } | |
| 92 | ||
| 93 | function Missing({ names, what }: { names: string[]; what: string }) { | |
| 94 | if (names.length === 0) return null; | |
| 95 | return ( | |
| 96 | <p className="flex items-center gap-1.5 border-t border-line px-5 py-2.5 text-xs text-faint"> | |
| 97 | <AlertTriangle size={12} className="shrink-0" /> | |
| 98 | {listOf(names)} didn't answer, so {what} isn't listed. | |
| 99 | </p> | |
| 100 | ); | |
| 101 | } | |
| 102 | ||
| 103 | // --- The window switch ---------------------------------------------------------- | |
| 104 | ||
| 105 | /** Since last visit, last 24 hours, last 7 days: links, so the span is in the address and works without script. */ | |
| 106 | export function WindowSwitch({ span }: { span: Pick<Span, "key"> }) { | |
| 107 | return ( | |
| 108 | <nav aria-label="What happened over" className="flex shrink-0 self-start rounded-lg border border-line bg-surface p-0.5"> | |
| 109 | {WINDOWS.map((option) => { | |
| 110 | const current = option.key === span.key; | |
| 111 | return ( | |
| 112 | <Link | |
| 113 | key={option.key} | |
| 114 | to={option.key === "last" ? "?" : `?window=${option.key}`} | |
| 115 | preventScrollReset | |
| 116 | aria-current={current ? "true" : undefined} | |
| 117 | className={cn( | |
| 118 | "rounded-md px-2.5 py-1 text-xs font-medium whitespace-nowrap transition-colors", | |
| 119 | current ? "bg-raised text-fg shadow-[0_0_0_1px_var(--color-line-strong)]" : "text-muted hover:text-fg", | |
| 120 | )} | |
| 121 | > | |
| 122 | {option.label} | |
| 123 | </Link> | |
| 124 | ); | |
| 125 | })} | |
| 126 | </nav> | |
| 127 | ); | |
| 128 | } | |
| 129 | ||
| 130 | // --- Where you're needed --------------------------------------------------------- | |
| 131 | ||
| 132 | const ICON: Record<AttentionKind, ReactNode> = { | |
| 133 | deploy: <Rocket size={16} className="text-danger" />, | |
| 134 | pull_blocked: <GitPullRequest size={16} className="text-warn" />, | |
| 135 | review: <GitPullRequest size={16} className="text-info" />, | |
| 136 | invitation: <Mail size={16} className="text-info" />, | |
| 137 | runner: <Server size={16} className="text-warn" />, | |
| 138 | notification: <Bell size={16} className="text-warn" />, | |
| 139 | chat: <AtSign size={16} className="text-accent" />, | |
| 140 | }; | |
| 141 | ||
| 142 | const STAKE_TONE: Record<NonNullable<Stake["tone"]> | "none", string> = { | |
| 143 | danger: "text-danger", | |
| 144 | warn: "text-warn", | |
| 145 | accent: "text-accent", | |
| 146 | none: "text-fg-soft", | |
| 147 | }; | |
| 148 | ||
| 149 | function StakeText({ stake }: { stake: Stake }) { | |
| 150 | return <span className={`text-xs font-medium tabular-nums ${STAKE_TONE[stake.tone ?? "none"]}`}>{stake.text}</span>; | |
| 151 | } | |
| 152 | ||
| 153 | /** Rows shown before the rest fold away. */ | |
| 154 | const ROWS = 8; | |
| 155 | ||
| 156 | function NeedItem({ row, now }: { row: AttentionRow; now: number }) { | |
| 157 | return ( | |
| 158 | <li className="flex items-start gap-3 border-t border-line px-5 py-3 first:border-t-0"> | |
| 159 | <span className="mt-0.5 shrink-0">{ICON[row.kind]}</span> | |
| 160 | <div className="min-w-0 grow"> | |
| 161 | <Link to={row.action.to} className="line-clamp-2 text-sm font-semibold text-fg hover:underline sm:block sm:truncate"> | |
| 162 | {row.title} | |
| 163 | </Link> | |
| 164 | <p className="truncate text-xs text-muted">{row.detail}</p> | |
| 165 | <div className="mt-1 flex flex-wrap items-center gap-x-2 text-xs sm:hidden"> | |
| 166 | {row.stake && <StakeText stake={row.stake} />} | |
| 167 | {row.at > 0 && ( | |
| 168 | <span className="text-faint" suppressHydrationWarning> | |
| 169 | waiting {waited(row.at, now)} | |
| 170 | </span> | |
| 171 | )} | |
| 172 | </div> | |
| 173 | </div> | |
| 174 | <div className="hidden max-w-48 shrink-0 pt-0.5 text-right whitespace-nowrap sm:block"> | |
| 175 | {row.stake && <StakeText stake={row.stake} />} | |
| 176 | {row.at > 0 && ( | |
| 177 | <div className="text-[0.6875rem] text-faint tabular-nums" suppressHydrationWarning> | |
| 178 | waiting {waited(row.at, now)} | |
| 179 | </div> | |
| 180 | )} | |
| 181 | </div> | |
| The site is built from its shared components: one Button with its variants and sizes, one Badge, Card for every panel, Alert for notices, Kbd for keys, Table for the security tables, and Avatar, Command, Progress, ScrollArea, ToggleGroup, Collapsible and Accordion in the kit; raw buttons, hand-drawn cards and pills and per-file button class lists are gone, outline buttons stand the same height as filled ones, and the phone's Ask g1t loses its tile. Every page was compared pixel by pixel before and after. | 182 | <ButtonLink to={row.action.to} variant="outline" size="xs" className="border-line-strong px-2.5 text-fg hover:border-accent/50 hover:bg-raised"> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 183 | {row.action.label} |
| The site is built from its shared components: one Button with its variants and sizes, one Badge, Card for every panel, Alert for notices, Kbd for keys, Table for the security tables, and Avatar, Command, Progress, ScrollArea, ToggleGroup, Collapsible and Accordion in the kit; raw buttons, hand-drawn cards and pills and per-file button class lists are gone, outline buttons stand the same height as filled ones, and the phone's Ask g1t loses its tile. Every page was compared pixel by pixel before and after. | 184 | </ButtonLink> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 185 | </li> |
| 186 | ); | |
| 187 | } | |
| 188 | ||
| 189 | /** Everything waiting on you, most pressing and longest waiting first; past the first few, the rest fold away. */ | |
| 190 | export function NeedsList({ attention, now }: { attention: Attention; now: number }) { | |
| 191 | const { rows, missing } = attention; | |
| 192 | return ( | |
| 193 | <> | |
| 194 | {rows.length === 0 ? ( | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 195 | <Quiet>{missing.length > 0 ? "Nothing needs you in what could be read." : "Nothing needs you. Reviews, invitations and messages for you show here as soon as they arrive."}</Quiet> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 196 | ) : ( |
| 197 | <ul> | |
| 198 | {rows.slice(0, ROWS).map((row) => ( | |
| 199 | <NeedItem key={row.key} row={row} now={now} /> | |
| 200 | ))} | |
| 201 | </ul> | |
| 202 | )} | |
| 203 | {rows.length > ROWS && ( | |
| 204 | <details className="group border-t border-line"> | |
| 205 | <summary className="cursor-pointer list-none px-5 py-2.5 text-xs font-medium text-fg-soft hover:text-fg"> | |
| 206 | <span className="group-open:hidden">Show {(rows.length - ROWS).toLocaleString("en-US")} more</span> | |
| 207 | <span className="hidden group-open:inline">Show fewer</span> | |
| 208 | </summary> | |
| 209 | <ul className="border-t border-line"> | |
| 210 | {rows.slice(ROWS).map((row) => ( | |
| 211 | <NeedItem key={row.key} row={row} now={now} /> | |
| 212 | ))} | |
| 213 | </ul> | |
| 214 | </details> | |
| 215 | )} | |
| 216 | <Missing names={missing} what="what waits there" /> | |
| 217 | </> | |
| 218 | ); | |
| 219 | } | |
| 220 | ||
| 221 | export function RowsSkeleton({ rows = 3 }: { rows?: number }) { | |
| 222 | return ( | |
| 223 | <div className="space-y-4 px-5 pb-5" aria-busy="true"> | |
| 224 | {Array.from({ length: rows }, (_, i) => ( | |
| 225 | <div key={i} className="flex items-center gap-3"> | |
| 226 | <Skeleton className="size-4 rounded" /> | |
| 227 | <div className="grow space-y-1.5"> | |
| 228 | <Skeleton className="h-3.5" style={{ width: `${60 - i * 8}%` }} /> | |
| 229 | <Skeleton className="h-3" style={{ width: `${80 - i * 10}%` }} /> | |
| 230 | </div> | |
| 231 | <Skeleton className="h-7 w-16" /> | |
| 232 | </div> | |
| 233 | ))} | |
| 234 | </div> | |
| 235 | ); | |
| 236 | } | |
| 237 | ||
| 238 | // --- Start here ----------------------------------------------------------------- | |
| 239 | ||
| 240 | export function StartHere({ attention, now }: { attention: Attention; now: number }) { | |
| 241 | const start = attention.start; | |
| 242 | return ( | |
| 243 | <Card | |
| 244 | title={ | |
| 245 | <> | |
| 246 | Start here <span className="font-normal text-faint">· picked by g1t</span> | |
| 247 | </> | |
| 248 | } | |
| 249 | > | |
| 250 | {!start ? ( | |
| 251 | <Quiet>{attention.missing.length > 0 ? "Nothing to pick from what could be read." : "Nothing needs you, so there's nothing to pick."}</Quiet> | |
| 252 | ) : ( | |
| 253 | <div className="px-5 pb-5"> | |
| 254 | <div className="flex items-start gap-3"> | |
| 255 | <span className="mt-0.5 shrink-0">{ICON[start.row.kind]}</span> | |
| 256 | <div className="min-w-0"> | |
| 257 | <p className="text-[0.9375rem] leading-snug font-semibold text-fg">{start.row.title}</p> | |
| 258 | <p className="mt-1 line-clamp-2 text-xs text-muted">{start.row.detail}</p> | |
| 259 | </div> | |
| 260 | </div> | |
| 261 | <dl className="mt-4 grid grid-cols-[auto_1fr] gap-x-4 gap-y-1.5 text-xs"> | |
| 262 | {start.row.owner && ( | |
| 263 | <> | |
| 264 | <dt className="text-muted">Owner</dt> | |
| 265 | <dd className="flex min-w-0 items-center gap-1.5 text-fg-soft"> | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 266 | <User size={12} className="shrink-0 text-muted" /> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 267 | <span className="truncate">{start.row.owner.agent ? `@${start.row.owner.name}` : start.row.owner.name}</span> |
| 268 | </dd> | |
| 269 | </> | |
| 270 | )} | |
| 271 | <dt className="text-muted">Next</dt> | |
| 272 | <dd className="text-fg-soft">{start.row.action.label}</dd> | |
| 273 | {start.row.at > 0 && ( | |
| 274 | <> | |
| 275 | <dt className="text-muted">Waiting</dt> | |
| 276 | <dd className="text-fg-soft" suppressHydrationWarning> | |
| 277 | {waited(start.row.at, now)} | |
| 278 | </dd> | |
| 279 | </> | |
| 280 | )} | |
| 281 | {start.row.stake && ( | |
| 282 | <> | |
| 283 | <dt className="text-muted">At stake</dt> | |
| 284 | <dd> | |
| 285 | <StakeText stake={start.row.stake} /> | |
| 286 | </dd> | |
| 287 | </> | |
| 288 | )} | |
| 289 | </dl> | |
| 290 | <p className="mt-4 rounded-lg bg-raised/60 px-3 py-2.5 text-xs leading-relaxed text-fg-soft"> | |
| 291 | <span className="font-medium text-fg">Why this one: </span> | |
| 292 | {start.why} | |
| 293 | </p> | |
| 294 | <Link | |
| 295 | to={start.row.action.to} | |
| 296 | className="mt-4 inline-flex h-8 items-center gap-1.5 rounded-lg bg-accent px-3 text-sm font-medium text-bg transition-colors hover:bg-accent-hover" | |
| 297 | > | |
| 298 | {start.row.action.label} | |
| 299 | <ArrowRight size={14} /> | |
| 300 | </Link> | |
| 301 | </div> | |
| 302 | )} | |
| 303 | </Card> | |
| 304 | ); | |
| 305 | } | |
| 306 | ||
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 307 | // --- What people did in the span ------------------------------------------------ |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 308 | |
| 309 | /** "in the span" said the way the span is: "since you were last here", "in the last 24 hours". */ | |
| 310 | function inSpan(span: Pick<Span, "key" | "note" | "words">): string { | |
| 311 | return span.key === "last" && span.note == null ? "since you were last here" : span.words.charAt(0).toLowerCase() + span.words.slice(1); | |
| 312 | } | |
| 313 | ||
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 314 | /** The faces of the most active, with a tooltip naming each, and how many more there were. */ |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 315 | const FACES = 5; |
| 316 | ||
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 317 | function Faces({ actors }: { actors: DigestActor[] }) { |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 318 | if (actors.length === 0) return null; |
| 319 | const shown = actors.slice(0, FACES); | |
| 320 | const more = actors.length - shown.length; | |
| 321 | return ( | |
| 322 | <div className="flex items-center gap-1.5"> | |
| 323 | <ul className="flex -space-x-1.5" aria-label="Most active"> | |
| 324 | {shown.map((actor) => ( | |
| 325 | <li key={actor.key} className="rounded-full ring-2 ring-surface"> | |
| 326 | <Tooltip> | |
| 327 | <TooltipTrigger asChild> | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 328 | <span className="block"> |
| 329 | <Avatar name={actor.name} size={22} /> | |
| 330 | </span> | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 331 | </TooltipTrigger> |
| 332 | <TooltipContent> | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 333 | {actor.name} · {actor.count.toLocaleString("en-US")} {actor.count === 1 ? "thing" : "things"} |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 334 | </TooltipContent> |
| 335 | </Tooltip> | |
| 336 | </li> | |
| 337 | ))} | |
| 338 | </ul> | |
| 339 | {more > 0 && <span className="text-xs text-muted">+{more.toLocaleString("en-US")}</span>} | |
| 340 | </div> | |
| 341 | ); | |
| 342 | } | |
| 343 | ||
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 344 | /** The lines: each what was done, with its count in the words, linked to where it is listed. */ |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 345 | function Lines({ column }: { column: DigestColumn }) { |
| 346 | return ( | |
| 347 | <ul className="mt-3 space-y-1.5"> | |
| 348 | {column.lines.map((line) => ( | |
| 349 | <li key={line.key} className="flex items-start gap-2 text-sm text-fg"> | |
| 350 | <span className="mt-1.75 size-1.5 shrink-0 rounded-full bg-fg-soft/60" aria-hidden /> | |
| 351 | {line.to ? ( | |
| 352 | <Link to={line.to} className="hover:underline"> | |
| 353 | {line.text} | |
| 354 | </Link> | |
| 355 | ) : ( | |
| 356 | <span>{line.text}</span> | |
| 357 | )} | |
| 358 | </li> | |
| 359 | ))} | |
| 360 | </ul> | |
| 361 | ); | |
| 362 | } | |
| 363 | ||
| 364 | /** | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 365 | * What people did in the span: lines with counts linked to where each |
| 366 | * thing is listed, and the most active faces. It says when it is empty, | |
| 367 | * and which part didn't answer. | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 368 | */ |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 369 | export function PeopleCard({ digest, span, code }: { digest: Digest | null; span: Pick<Span, "key" | "note" | "words">; code: boolean }) { |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 370 | const when = inSpan(span); |
| 371 | if (!digest) { | |
| 372 | return ( | |
| 373 | <section className="rounded-2xl border border-line bg-surface p-5"> | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 374 | <p className="text-sm text-muted">What happened couldn't be read right now. It shows here once Code and Chat answer.</p> |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 375 | </section> |
| 376 | ); | |
| 377 | } | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 378 | const silent = digest.missing.filter((name) => name === "Code" || name === "Chat"); |
| 379 | const col = digest.people; | |
| 380 | return ( | |
| 381 | <section className="rounded-2xl border border-line bg-surface p-5" aria-label="What people did"> | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 382 | <div className="flex flex-wrap items-center gap-3"> |
| 383 | <h3 className="inline-flex items-center gap-1.5 text-sm font-semibold text-fg"> | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 384 | <Users size={15} className="text-muted" /> |
| 385 | People | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 386 | </h3> |
| 387 | {col.actors.length > 0 && ( | |
| 388 | <Badge tone="neutral" size="md" className="tabular-nums"> | |
| 389 | {col.actors.length.toLocaleString("en-US")} active | |
| 390 | </Badge> | |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 391 | )} |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 392 | <span className="ml-auto"> |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 393 | <Faces actors={col.actors} /> |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 394 | </span> |
| 395 | </div> | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 396 | {col.lines.length > 0 ? ( |
| 397 | <Lines column={col} /> | |
| 398 | ) : ( | |
| 399 | <p className="mt-3 text-sm text-muted"> | |
| 400 | {silent.length > 0 | |
| 401 | ? `Nothing from what could be read ${when}.` | |
| 402 | : code | |
| 403 | ? `Nobody pushed, opened, merged, reviewed, edited a doc or said anything ${when}.` | |
| 404 | : `Nobody edited a doc or said anything ${when}.`} | |
| 405 | </p> | |
| 406 | )} | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 407 | {silent.length > 0 && ( |
| 408 | <p className="mt-3 flex items-center gap-1.5 text-xs text-faint"> | |
| 409 | <AlertTriangle size={12} className="shrink-0" /> | |
| 410 | {listOf(silent)} didn't answer, so {silent.length === 1 ? "its part" : "their parts"} {silent.length === 1 ? "isn't" : "aren't"} counted. | |
| 411 | </p> | |
| 412 | )} | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 413 | {digest.partial && <p className="mt-3 text-xs text-faint">More happened than Home reads at once, so some counts are low.</p>} |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 414 | </section> |
| 415 | ); | |
| 416 | } | |
| 417 | ||
| 418 | export function SpanSkeleton() { | |
| 419 | return ( | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 420 | <section className="space-y-3 rounded-2xl border border-line bg-surface p-5" aria-busy="true"> |
| 421 | <div className="flex items-center justify-between"> | |
| 422 | <Skeleton className="h-3.5 w-20" /> | |
| 423 | <Skeleton className="h-5 w-24 rounded-full" /> | |
| 424 | </div> | |
| 425 | <Skeleton className="h-3.5 w-3/4" /> | |
| 426 | <Skeleton className="h-3.5 w-1/2" /> | |
| 427 | <Skeleton className="h-3.5 w-2/3" /> | |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 428 | </section> |
| 429 | ); | |
| 430 | } | |
| 431 | ||
| 432 | // --- Landed, builds and decisions ------------------------------------------------- | |
| 433 | ||
| 434 | /** Rows each list shows before the rest fold away. */ | |
| 435 | const LIST_ROWS = 5; | |
| 436 | ||
| 437 | function When({ at, now, tz }: { at: number; now: number; tz: string | null }) { | |
| 438 | return ( | |
| 439 | <span className="shrink-0 text-xs text-faint tabular-nums" suppressHydrationWarning> | |
| 440 | {whenShort(at, now, tz)} | |
| 441 | </span> | |
| 442 | ); | |
| 443 | } | |
| 444 | ||
| 445 | function Folded<T>({ items, render }: { items: T[]; render: (item: T) => ReactNode }) { | |
| 446 | return ( | |
| 447 | <> | |
| 448 | <ul>{items.slice(0, LIST_ROWS).map(render)}</ul> | |
| 449 | {items.length > LIST_ROWS && ( | |
| 450 | <details className="group border-t border-line"> | |
| 451 | <summary className="cursor-pointer list-none px-5 py-2.5 text-xs font-medium text-fg-soft hover:text-fg"> | |
| 452 | <span className="group-open:hidden">Show {(items.length - LIST_ROWS).toLocaleString("en-US")} more</span> | |
| 453 | <span className="hidden group-open:inline">Show fewer</span> | |
| 454 | </summary> | |
| 455 | <ul className="border-t border-line">{items.slice(LIST_ROWS).map(render)}</ul> | |
| 456 | </details> | |
| 457 | )} | |
| 458 | </> | |
| 459 | ); | |
| 460 | } | |
| 461 | ||
| 462 | const rowClass = "flex items-center gap-3 border-t border-line px-5 py-2.5 first:border-t-0"; | |
| 463 | ||
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 464 | const LANDED_ICON: Record<LandedChange["kind"], ReactNode> = { |
| 465 | merge: <GitMerge size={15} className="shrink-0 text-success" />, | |
| 466 | push: <GitCommitHorizontal size={15} className="shrink-0 text-success" />, | |
| 467 | deploy: <Rocket size={15} className="shrink-0 text-success" />, | |
| 468 | release: <Tag size={15} className="shrink-0 text-info" />, | |
| 469 | package: <Package size={15} className="shrink-0 text-info" />, | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 470 | }; |
| 471 | ||
| 472 | /** A few words on the card's line: how many of each kind landed. */ | |
| 473 | function landedHint(landed: LandedChange[]): string { | |
| 474 | const count = (kind: LandedChange["kind"]) => landed.filter((change) => change.kind === kind).length; | |
| 475 | const parts: string[] = []; | |
| 476 | const merged = count("merge"); | |
| 477 | const pushes = count("push"); | |
| 478 | const deploys = count("deploy"); | |
| 479 | const published = count("release") + count("package"); | |
| 480 | if (merged > 0) parts.push(`${merged.toLocaleString("en-US")} merged`); | |
| 481 | if (pushes > 0) parts.push(`${pushes.toLocaleString("en-US")} pushed direct`); | |
| 482 | if (deploys > 0) parts.push(`${deploys.toLocaleString("en-US")} ${deploys === 1 ? "deploy" : "deploys"}`); | |
| 483 | if (published > 0) parts.push(`${published.toLocaleString("en-US")} published`); | |
| 484 | return parts.join(" · "); | |
| 485 | } | |
| 486 | ||
| 487 | /** | |
| 488 | * Everything that landed in the span, newest first: pull requests merged, | |
| 489 | * commits pushed straight to a default branch, production builds that went | |
| 490 | * live, releases and package versions published. | |
| 491 | */ | |
| 492 | export function LandedCard({ landed, missing, code, now, tz }: { landed: LandedChange[] | null; missing: string[]; code: boolean; now: number; tz: string | null }) { | |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 493 | if (!code) return null; |
| 494 | return ( | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 495 | <Card title="Landed" hint={landed && landed.length > 0 ? landedHint(landed) : undefined}> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 496 | {!landed ? ( |
| 497 | <Quiet>Code didn't answer, so what landed can't be listed right now.</Quiet> | |
| 498 | ) : landed.length === 0 ? ( | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 499 | <Quiet>{missing.length > 0 ? "Nothing landed in what could be read." : "Nothing merged, pushed to a default branch, deployed to production or published in this span."}</Quiet> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 500 | ) : ( |
| 501 | <Folded | |
| 502 | items={landed} | |
| 503 | render={(change) => ( | |
| 504 | <li key={change.key} className={rowClass}> | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 505 | {LANDED_ICON[change.kind]} |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 506 | <div className="min-w-0 grow"> |
| 507 | <Link to={change.to} className="block truncate text-sm text-fg hover:underline"> | |
| 508 | {change.title} | |
| 509 | </Link> | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 510 | <p className="truncate text-xs text-muted">{change.detail}</p> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 511 | </div> |
| 512 | <When at={change.at} now={now} tz={tz} /> | |
| 513 | </li> | |
| 514 | )} | |
| 515 | /> | |
| 516 | )} | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 517 | {landed && <Missing names={missing} what="everything that landed" />} |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 518 | </Card> |
| 519 | ); | |
| 520 | } | |
| 521 | ||
| 522 | export function DeploysCard({ deploys, complete, code, now, tz }: { deploys: DeployRow[] | null; complete: boolean; code: boolean; now: number; tz: string | null }) { | |
| 523 | // A workspace that doesn't deploy has nothing to say here. | |
| 524 | if (!code || (deploys && deploys.length === 0)) return null; | |
| 525 | const failed = deploys?.filter((row) => row.outcome === "failed").length ?? 0; | |
| 526 | return ( | |
| 527 | <Card title="Deploys" hint={deploys ? `${deploys.length.toLocaleString("en-US")} finished${failed ? ` · ${failed} failed` : ""}` : undefined}> | |
| 528 | {!deploys ? ( | |
| 529 | <Quiet>Deployments didn't answer, so builds can't be listed right now.</Quiet> | |
| 530 | ) : ( | |
| 531 | <Folded | |
| 532 | items={deploys} | |
| 533 | render={(row) => ( | |
| 534 | <li key={row.key} className={rowClass}> | |
| 535 | <Rocket size={15} className={cn("shrink-0", row.outcome === "failed" ? "text-danger" : "text-success")} /> | |
| 536 | <div className="min-w-0 grow"> | |
| 537 | <Link to={row.to} className="block truncate text-sm text-fg hover:underline"> | |
| 538 | {row.kind === "production" ? "Production" : `Preview of ${row.branch ?? "a branch"}`} of {row.project} | |
| 539 | <span className={cn("ml-2 text-xs font-medium", row.outcome === "failed" ? "text-danger" : "text-success")}> | |
| 540 | {row.outcome === "failed" ? "Failed" : "Went live"} | |
| 541 | </span> | |
| 542 | </Link> | |
| 543 | <p className="truncate text-xs text-muted"> | |
| 544 | <span className="font-mono">{row.commit}</span> · by {row.by} | |
| 545 | </p> | |
| 546 | </div> | |
| 547 | <When at={row.at} now={now} tz={tz} /> | |
| 548 | </li> | |
| 549 | )} | |
| 550 | /> | |
| 551 | )} | |
| 552 | {deploys && !complete && <p className="border-t border-line px-5 py-2.5 text-xs text-faint">Some projects' builds couldn't be read.</p>} | |
| 553 | </Card> | |
| 554 | ); | |
| 555 | } | |
| 556 | ||
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 557 | export function DecisionsCard({ decisions, now, tz }: { decisions: DecisionRow[] | null; now: number; tz: string | null }) { |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 558 | if (!decisions || decisions.length === 0) return null; |
| 559 | return ( | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 560 | <Card title="Decisions" hint="Recorded in the workspace's memory"> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 561 | <Folded |
| 562 | items={decisions} | |
| 563 | render={(row) => ( | |
| 564 | <li key={row.key} className={rowClass}> | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 565 | <Lightbulb size={15} className="shrink-0 text-warn" /> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 566 | <div className="min-w-0 grow"> |
| 567 | <Link to={row.to} className="line-clamp-2 text-sm text-fg hover:underline"> | |
| 568 | {row.text} | |
| 569 | </Link> | |
| 570 | {row.by && <p className="truncate text-xs text-muted">by {row.by}</p>} | |
| 571 | </div> | |
| 572 | <When at={row.at} now={now} tz={tz} /> | |
| 573 | </li> | |
| 574 | )} | |
| 575 | /> | |
| 576 | </Card> | |
| 577 | ); | |
| 578 | } | |
| 579 | ||
| 580 | // --- Spend in the span ------------------------------------------------------------ | |
| 581 | ||
| 582 | /** | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 583 | * Each kind of charge keeps its colour, whatever its rank: sandboxes and |
| 584 | * runners one, hosting another, storage and the rest grey. | |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 585 | */ |
| 586 | function kindColour(kind: string): string { | |
| 587 | if (/sandbox|runner/i.test(kind)) return "bg-info"; | |
| 588 | if (/deploy|domain/i.test(kind)) return "bg-fg-soft"; | |
| 589 | if (/storage|git|cache|package/i.test(kind)) return "bg-muted"; | |
| 590 | return "bg-faint"; | |
| 591 | } | |
| 592 | ||
| 593 | const utcDay = (day: string) => new Intl.DateTimeFormat("en-US", { month: "short", day: "numeric", timeZone: "UTC" }).format(Date.parse(`${day}T12:00:00Z`)); | |
| 594 | ||
| 595 | export function SpendCard({ spend, slug, span }: { spend: Spend | null; slug: string; span: Pick<Span, "key" | "note" | "words"> }) { | |
| 596 | return ( | |
| 597 | <Card title="Spent" link={{ label: "Billing", to: `/${slug}/-/billing` }}> | |
| 598 | {!spend ? ( | |
| 599 | <Quiet>Billing didn't answer, so spend can't be shown right now.</Quiet> | |
| 600 | ) : ( | |
| 601 | <div className="px-5 pb-5"> | |
| Money is written one way. A single formatter turns millionths of a dollar into dollars, rounding half up on whole micros rather than on a float, so the same sum reads the same on every page: under a cent reads <$0.01 on a total and exactly nothing is $0.00, while the statement's lines, a session's receipt, the price book and an agent's effort costs carry up to four places where the fraction of a cent is the point; the six formatters that each rounded their own way are gone. This month is the calendar month in UTC from its first day to today everywhere, and billing counts the month's not-yet-closed usage in any range that reaches into the current month, so the top bar's pill, Spend, Home and Usage ask for the same days and get the same figure; Home now reads the usage report the others read instead of adding up statements. Usage's pending sentence says what of that usage the close will charge after the discount and included usage, which the billing API returns as pending_charged_micros. A reconciliation test holds the pill, Spend, Home and Usage to one number for one month. The usage and billing guide says how amounts are written and what this month means. | 602 | <div className="text-3xl font-semibold tracking-tight text-fg tabular-nums">{money(spend.totalMicros)}</div> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 603 | <p className="mt-1 text-xs text-muted"> |
| Money is written one way. A single formatter turns millionths of a dollar into dollars, rounding half up on whole micros rather than on a float, so the same sum reads the same on every page: under a cent reads <$0.01 on a total and exactly nothing is $0.00, while the statement's lines, a session's receipt, the price book and an agent's effort costs carry up to four places where the fraction of a cent is the point; the six formatters that each rounded their own way are gone. This month is the calendar month in UTC from its first day to today everywhere, and billing counts the month's not-yet-closed usage in any range that reaches into the current month, so the top bar's pill, Spend, Home and Usage ask for the same days and get the same figure; Home now reads the usage report the others read instead of adding up statements. Usage's pending sentence says what of that usage the close will charge after the discount and included usage, which the billing API returns as pending_charged_micros. A reconciliation test holds the pill, Spend, Home and Usage to one number for one month. The usage and billing guide says how amounts are written and what this month means. | 604 | {spend.from === spend.to ? utcDay(spend.from) : `${utcDay(spend.from)} – ${utcDay(spend.to)}`} |
| 605 | {spend.monthMicros != null && ` · ${money(spend.monthMicros)} this month so far`} | |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 606 | </p> |
| 607 | {spend.lines.length === 0 ? ( | |
| 608 | <p className="mt-4 text-sm text-muted">Nothing on the statement {inSpan(span)}.</p> | |
| 609 | ) : ( | |
| 610 | <> | |
| 611 | <div className="mt-4 flex h-2.5 gap-[2px] overflow-hidden rounded-full" role="img" aria-label="Spend by kind of charge"> | |
| 612 | {spend.lines.map((line) => ( | |
| 613 | <span key={line.kind} className={`h-full min-w-[3px] ${kindColour(line.kind)}`} style={{ flexGrow: line.micros }} /> | |
| 614 | ))} | |
| 615 | </div> | |
| 616 | <ul className="mt-3 space-y-2"> | |
| 617 | {spend.lines.map((line) => ( | |
| 618 | <li key={line.kind} className="flex items-center gap-2 text-sm"> | |
| 619 | <span className={`size-2 shrink-0 rounded-[2px] ${kindColour(line.kind)}`} aria-hidden /> | |
| 620 | <span className="min-w-0 grow truncate text-fg-soft">{line.kind}</span> | |
| Money is written one way. A single formatter turns millionths of a dollar into dollars, rounding half up on whole micros rather than on a float, so the same sum reads the same on every page: under a cent reads <$0.01 on a total and exactly nothing is $0.00, while the statement's lines, a session's receipt, the price book and an agent's effort costs carry up to four places where the fraction of a cent is the point; the six formatters that each rounded their own way are gone. This month is the calendar month in UTC from its first day to today everywhere, and billing counts the month's not-yet-closed usage in any range that reaches into the current month, so the top bar's pill, Spend, Home and Usage ask for the same days and get the same figure; Home now reads the usage report the others read instead of adding up statements. Usage's pending sentence says what of that usage the close will charge after the discount and included usage, which the billing API returns as pending_charged_micros. A reconciliation test holds the pill, Spend, Home and Usage to one number for one month. The usage and billing guide says how amounts are written and what this month means. | 621 | <span className="text-fg tabular-nums">{money(line.micros)}</span> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 622 | </li> |
| 623 | ))} | |
| 624 | </ul> | |
| 625 | </> | |
| 626 | )} | |
| 627 | <p className="mt-4 text-[0.6875rem] leading-relaxed text-faint"> | |
| 628 | At price, as your statement shows it. The statement keeps whole days, midnight to midnight UTC, so this counts from the start of the day the span began. | |
| 629 | </p> | |
| 630 | </div> | |
| 631 | )} | |
| 632 | </Card> | |
| 633 | ); | |
| 634 | } | |
| 635 | ||
| 636 | export function SpendSkeleton() { | |
| 637 | return ( | |
| 638 | <div className="space-y-3 px-5 pb-5" aria-busy="true"> | |
| 639 | <Skeleton className="h-8 w-28" /> | |
| 640 | <Skeleton className="h-2.5 w-full rounded-full" /> | |
| 641 | <Skeleton className="h-3 w-3/4" /> | |
| 642 | <Skeleton className="h-3 w-2/3" /> | |
| 643 | </div> | |
| 644 | ); | |
| 645 | } | |
| 646 | ||
| 647 | // --- Running now ------------------------------------------------------------------- | |
| 648 | ||
| 649 | const RUNNING_ICON: Record<RunningRow["kind"], ReactNode> = { | |
| Home says what people did as well as what agents did. Since you were last here now has a People column and an Agents column: who pushed how many commits to which projects, pull requests opened, merged and reviewed, issues opened and closed, docs edited, messages sent and deploys that went out, each line a link to where those are listed, with the agents' acceptance (first time, after review, didn't finish) kept as a row of their column; the sentence under the heading sums it up honestly, and says when a part of g1t could not be read, or that the span was quiet. Landed counts merged pull requests, commits pushed straight to a default branch, production deploys that went live, releases and packages, newest first; Running now adds workflow runs. Behind it, the events service answers an activity digest over a span in one round trip from its existing indexes, every push now records how many commits it carried, and chat counts the messages sent in the conversations you can read. The Home guide defines every line, and says how this scales. | 650 | workflow: <Workflow size={15} className="text-info" />, |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 651 | deploy: <Rocket size={15} className="text-info" />, |
| 652 | }; | |
| 653 | ||
| 654 | export function RunningList({ rows, missing, now }: { rows: RunningRow[]; missing: string[]; now: number }) { | |
| 655 | return ( | |
| 656 | <> | |
| 657 | {rows.length === 0 ? ( | |
| The site no longer has g1t's own agents. g1t is where the coding agents you already use do their work, so the Agents mode is gone from the rail, with every page under /<workspace>/-/agents: personas, faces, skills, abilities, routines, computers, sessions and agent spend. Customize (skills, connectors, plugins, marketplaces), the context hub, model providers, the AI Gateway page, the Spend page and pill, Ask g1t, and the repository's Agents, Plans and agent settings pages go too, and their addresses now answer 404. | 658 | <Quiet>{missing.length > 0 ? "Nothing is running in what could be read." : "Nothing is running. Workflow runs and builds show here while they go."}</Quiet> |
| The front page is Home: where you're needed comes first, however long ago it started, with what's at stake and how long it has waited; then what happened since you were last here, in words, with the agent work that settled, what landed, what deployed, what was decided and what was spent, switchable to the last 24 hours or 7 days; then what's running now. Your last visit is kept with your account, so it's the same on every device; Today's old address leads here. The home guide says how. | 659 | ) : ( |
| 660 | <Folded | |
| 661 | items={rows} | |
| 662 | render={(row) => ( | |
| 663 | <li key={row.key} className={rowClass}> | |
| 664 | <span className="shrink-0">{RUNNING_ICON[row.kind]}</span> | |
| 665 | <div className="min-w-0 grow"> | |
| 666 | <Link to={row.to} className="block truncate text-sm text-fg hover:underline"> | |
| 667 | {row.title} | |
| 668 | </Link> | |
| 669 | <p className="truncate text-xs text-muted">{row.detail}</p> | |
| 670 | </div> | |
| 671 | <span className="inline-flex shrink-0 items-center gap-1.5 rounded-full bg-raised px-2 py-0.5 text-xs text-fg-soft"> | |
| 672 | <Loader size={11} className="animate-spin text-accent motion-reduce:animate-none" aria-hidden /> | |
| 673 | {row.status} | |
| 674 | </span> | |
| 675 | <span className="hidden w-14 shrink-0 text-right text-xs text-faint tabular-nums sm:inline" suppressHydrationWarning> | |
| 676 | {waited(row.at, now)} | |
| 677 | </span> | |
| 678 | </li> | |
| 679 | )} | |
| 680 | /> | |
| 681 | )} | |
| 682 | <Missing names={missing} what="everything running" /> | |
| 683 | </> | |
| 684 | ); | |
| 685 | } |
