Self-hosted, Self-Improving, Federated Everything App — dev management, agent orchestration, persistent mind, digital twin, second brain, and creative pipeline from a single dashboard over Tailscale to port the user into the OS
JavaScript ai-agentsdashboarddeveloper-toolsdigital-twinquantified-selfsecond-brainself-hostedtailscale
import { useState } from 'react';
import { Lightbulb, RefreshCw, X } from 'lucide-react';
import * as api from '../../services/api';
import toast from '../ui/Toast';
import BrailleSpinner from '../BrailleSpinner';
import RepoStudyFields from './RepoStudyFields';
import { useAsyncAction } from '../../hooks/useAsyncAction.js';
import { useRepoStudyConfig } from '../../hooks/useRepoStudyConfig.js';
/**
* The Links tab's on-demand "Update & study" form for an already-cloned repo:
* refresh the checkout, then queue a fresh `repo-study` run with a brief the
* user writes right now ("look at how it does X", "compare its onboarding to
* ours") instead of the one they may have written weeks ago at capture time.
*
* The knobs and their payload come from the SAME `useRepoStudyConfig` /
* `RepoStudyFields` pair the capture boxes use, so the two dispatch paths can't
* drift apart. The brief the last run was given seeds the field.
*
* @param {object} props
* @param {object} props.link the cloned repo link
* @param {() => void} props.onClose
* @param {(link: object) => void} props.onQueued receives the updated link record
*/
export default function RepoRestudyPanel({ link, onClose, onQueued }) {
const study = useRepoStudyConfig({
resetKey: link.id,
initialStudyContext: link.repoStudy?.studyContext || '',
});
const [pull, setPull] = useState(true);
const [handleSubmit, submitting] = useAsyncAction(async () => {
const result = await api.studyBrainLink(link.id, { pull, ...study.studyPayload() }, { silent: true });
// A pull failure does not stop the study — say so, or the user has no way to
// know the agent is reading a checkout that never got refreshed.
const pullFailed = result.pulled?.ok === false;
toast[pullFailed ? 'warning' : 'success'](pullFailed
? 'Study queued — the pull failed, so it reads the existing checkout'
: 'Study queued — track it in Chief of Staff');
onQueued?.(result.link);
onClose?.();
}, { errorMessage: 'Failed to queue study' });
return (
<div className="mt-2 w-full rounded-lg border border-port-border bg-port-bg/60 p-3 space-y-2">
<div className="flex items-center gap-2">
<Lightbulb size={14} className="text-port-accent shrink-0" />
<h4 className="text-sm font-medium text-gray-200">Update & study this repo</h4>
<button
onClick={onClose}
className="ml-auto flex min-h-[44px] min-w-[44px] items-center justify-center text-gray-400 hover:text-white transition-colors"
title="Close" aria-label="Close study form"
>
<X size={14} />
</button>
</div>
<label htmlFor={`restudy-pull-${link.id}`} className="flex items-center gap-2 text-xs text-gray-400">
<input
id={`restudy-pull-${link.id}`}
type="checkbox"
checked={pull}
onChange={e => setPull(e.target.checked)}
className="accent-port-accent"
/>
Pull the latest commits before studying
</label>
<RepoStudyFields
idPrefix={`restudy-${link.id}`}
{...study}
contextLabel="What should this study look for?"
contextPlaceholder="e.g. how it handles offline sync, and whether our Brain could adopt the same conflict model"
/>
<button
onClick={handleSubmit}
disabled={submitting}
className="flex items-center gap-1.5 px-3 py-1.5 text-xs rounded bg-port-accent/20 text-port-accent border border-port-accent/30 hover:bg-port-accent/30 transition-colors disabled:opacity-50 disabled:cursor-not-allowed"
>
{submitting ? <BrailleSpinner /> : <RefreshCw size={12} />}
{pull ? 'Update & study' : 'Study now'}
</button>
</div>
);
}