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, useEffect, useCallback, useMemo, useId } from 'react';
import { Loader2, Terminal, Wand2 } from 'lucide-react';
import Drawer from '../Drawer';
import ProviderModelSelector from '../ProviderModelSelector';
import ReviewerPicker from '../cos/ReviewerPicker';
import { DEFAULT_PR_COMPLETION, DEPFREE_MODE_DEFAULT, DEPFREE_MODE_OPTIONS, PR_COMPLETION_OPTIONS, depfreeModeOption, prCompletionOption } from '../cos/constants';
import useRunWithPicker from '../../hooks/useRunWithPicker';
import useReviewerModelOptions from '../../hooks/useReviewerModelOptions';
import { reviewerModelsFromDefaults, reviewerEffortsFromDefaults } from '../../lib/reviewerModels';
import { CodeReviewDefaultsProvider, useCodeReviewDefaults } from '../../hooks/useCodeReviewDefaults';
import useClaimReviewers from '../../hooks/useClaimReviewers';
import ClaimReviewerSource from './ClaimReviewerSource';
import WorkItemPicker from './WorkItemPicker';
import * as api from '../../services/api';
/**
* Pre-flight settings for an Agent Operations `/do:*` run: for `/do:next`, which
* work item to claim (or let the agent decide); for every command, the provider /
* model / effort pin and the task options (simplify, review-loop reviewers).
*
* Every control is optional — submitting untouched queues exactly the run the
* bare button used to. Reviewer fields are sent ONLY when the user edits them, so
* an untouched drawer still resolves the app's configured claim-work reviewers
* server-side rather than pinning whatever this form happened to display.
* Untouched completion settings also remain inherited: a slow preview must not
* turn the displayed fallback into an override of an app's draft/leave-open
* policy. The server pins the resolved policy when it queues the task.
*
* Mount only while open (the parent conditionally renders it): the provider fetch
* runs on mount unless the overview supplies its shared agent picker. Task
* settings reset on unmount; a supplied picker retains the overview overrides.
*/
function SlashDoRunDrawerBody({ open, command, label, appId, appName, onClose, onQueued, agentPicker }) {
const codeReviewDefaults = useCodeReviewDefaults();
// What a claim actually resolves for this app — the claim-work override layer
// the defaults above cannot see. Only `/do:next` reads reviewers server-side,
// so no other command pays for the lookup. `installed` still comes from the
// defaults, which is why both are fetched.
const claimReviewers = useClaimReviewers(command === 'next' ? appId : null);
// Resolved model lists for the reviewer table's Model column (the picker never
// fetches — see its `modelOptions` prop).
const reviewerModelOptions = useReviewerModelOptions();
const localPicker = useRunWithPicker();
const picker = agentPicker ?? localPicker;
const [simplify, setSimplify] = useState(true);
// Seeded from what the RUN will resolve for display. `review` staying null is
// what gates whether the fields are SENT — see the component doc.
const [review, setReview] = useState(null);
// Seeded display for an untouched picker. It has to show the reviewers the RUN
// would resolve, which is not the Code Review Defaults whenever a claim-work
// override is in play (see `GET /apps/:id/claim-reviewers`). The defaults are
// the fallback for the window before the lookup lands and for one that failed —
// they carry per-reviewer pins as `<reviewer>Model` / `<reviewer>Effort`
// scalars, which the picker takes as token-keyed maps.
const seededReview = useMemo(
() => claimReviewers ?? {
...codeReviewDefaults,
reviewerModels: reviewerModelsFromDefaults(codeReviewDefaults),
reviewerEfforts: reviewerEffortsFromDefaults(codeReviewDefaults),
},
[claimReviewers, codeReviewDefaults]
);
const reviewValue = review ?? seededReview;
const prCompletionId = useId();
const [prCompletion, setPrCompletion] = useState(DEFAULT_PR_COMPLETION);
const [prCompletionTouched, setPrCompletionTouched] = useState(false);
useEffect(() => {
if (!prCompletionTouched && claimReviewers?.prCompletion) {
setPrCompletion(claimReviewers.prCompletion);
}
}, [claimReviewers?.prCompletion, prCompletionTouched]);
const [work, setWork] = useState({ mode: 'auto', target: '', issueAuthorFilter: '' });
const [submitting, setSubmitting] = useState(false);
const [submitError, setSubmitError] = useState('');
const handleWorkChange = useCallback((next) => setWork(next), []);
const isNext = command === 'next';
// In pick mode an unselected item must block, not silently fall back to the
// agent-picked run the user just opted out of.
const awaitingPick = isNext && work.mode === 'pick' && !work.target;
// `/do:depfree`-only: the removal bar the route forwards as `--heavy` /
// `--extreme`. Sent only for that command — no other body reads the flag.
const isDepfree = command === 'depfree';
const depfreeModeId = useId();
const [depfreeMode, setDepfreeMode] = useState(DEPFREE_MODE_DEFAULT);
const handleQueue = async () => {
setSubmitting(true);
setSubmitError('');
// silent: this drawer owns its own inline error + the caller's toast.
const result = await api.createSlashdoTask(command, appId, {
target: work.target || undefined,
// Only send the filter/reviewers the user actually chose — otherwise the
// app's configured claim-work defaults apply server-side, unchanged.
issueAuthorFilter: work.issueAuthorFilter || undefined,
...picker.pin,
simplify,
...(isDepfree ? { depfreeMode } : {}),
// `isNext &&` states the route's contract rather than leaning on the picker
// being unrendered: only the `next` branch of POST /tasks/slashdo reads these,
// so sending them for another command would be a silent no-op.
...(isNext ? {
...(prCompletionTouched ? { prCompletion } : {}),
...(prCompletion === 'review-then-merge' && review ? {
reviewers: review.reviewers,
usernames: review.usernames,
optionalReviewers: review.optionalReviewers,
reviewerMaxRounds: review.reviewerMaxRounds,
reviewerModels: review.reviewerModels,
reviewerEfforts: review.reviewerEfforts
} : {})
} : {})
}, { silent: true }).catch((err) => {
setSubmitError(err.message || 'Failed to queue the task');
return null;
});
setSubmitting(false);
if (result) onQueued?.(result);
};
return (
<Drawer
open={open}
onClose={onClose}
title={`Run ${label}`}
subtitle={appName}
size="lg"
closeOnBackdrop={false}
>
<div className="space-y-6">
{isNext && <WorkItemPicker appId={appId} target={work.target} onChange={handleWorkChange} />}
<section className="space-y-3">
<div className="text-xs text-gray-500 uppercase tracking-wide">Agent</div>
<ProviderModelSelector {...picker.selectorProps} />
</section>
<section className="space-y-3">
<div className="text-xs text-gray-500 uppercase tracking-wide">Task settings</div>
<label className="flex items-center gap-2 cursor-pointer select-none">
<input
type="checkbox"
checked={simplify}
onChange={e => setSimplify(e.target.checked)}
className="w-4 h-4 rounded border-port-border bg-port-bg text-port-accent focus:ring-port-accent focus:ring-offset-0"
/>
<span className="flex items-center gap-1.5 text-sm text-gray-400">
<Wand2 size={14} className="text-port-accent-2" />
Simplify before committing
</span>
</label>
{isDepfree && (
<div>
<label htmlFor={depfreeModeId} className="block text-xs text-gray-400 mb-1">
Removal bar
</label>
<select
id={depfreeModeId}
value={depfreeMode}
onChange={e => setDepfreeMode(e.target.value)}
className="w-full bg-port-bg border border-port-border rounded px-3 py-2 text-white text-sm focus:border-port-accent focus:outline-hidden"
>
{DEPFREE_MODE_OPTIONS.map(option => (
<option key={option.value} value={option.value}>{option.label}</option>
))}
</select>
<p className="text-xs text-gray-500 mt-1">
{depfreeModeOption(depfreeMode)?.description}.
{depfreeMode === 'extreme' && ' Pick the strongest model and max effort above; the command puts that tier on every sub-agent it spawns.'}
</p>
</div>
)}
{/* Reviewer choices are `/do:next`-only: they're substituted into the
claim prompt's reviewer CSV, and `POST /tasks/slashdo` reads them only
on that branch. Every other `/do:*` body owns its own review/PR
sequence, so rendering the picker there would be four knobs wired to
nothing — the user would pick a reviewer and a model and the run would
silently discard both. */}
{isNext && (
<div>
<label htmlFor={prCompletionId} className="block text-xs text-gray-400 mb-1">
After opening PR
</label>
<select
id={prCompletionId}
aria-label="After opening PR"
value={prCompletion}
onChange={e => { setPrCompletion(e.target.value); setPrCompletionTouched(true); }}
className="w-full bg-port-bg border border-port-border rounded px-3 py-2 text-white text-sm focus:border-port-accent focus:outline-hidden"
>
{PR_COMPLETION_OPTIONS.map(option => (
<option key={option.value} value={option.value}>{option.label}</option>
))}
</select>
<p className="text-xs text-gray-500 mt-1">
{prCompletionOption(prCompletion)?.description}
</p>
</div>
)}
{isNext && prCompletion === 'review-then-merge' && (
<>
<ReviewerPicker
reviewers={reviewValue.reviewers}
usernames={reviewValue.usernames}
optionalReviewers={reviewValue.optionalReviewers}
reviewerMaxRounds={reviewValue.reviewerMaxRounds}
reviewerModels={reviewValue.reviewerModels}
reviewerEfforts={reviewValue.reviewerEfforts}
modelOptions={reviewerModelOptions}
installed={codeReviewDefaults.installed}
providerReviewUnsupported={codeReviewDefaults.providerReviewUnsupported}
// The claim flows substitute a reviewer CSV into their prompt and have
// no slashdo flag string, so stop-mode / reviewer-applies can't be honored.
showRunFlags={false}
onChange={({ reviewers, usernames, optionalReviewers, reviewerMaxRounds, reviewerModels, reviewerEfforts }) =>
setReview({ reviewers, usernames, optionalReviewers, reviewerMaxRounds, reviewerModels, reviewerEfforts })}
/>
<p className="text-xs text-gray-500">
The claim flow opens and merges its own PR, so these reviewers gate that merge (slashdo <code>--review-with</code>).
{!review && ' Leave them untouched to use this app’s configured reviewers.'}
</p>
{/* Which layer supplied the seeded list. A claim-work override wins
over Models → Code Reviewers silently, so a user who changed the
install default and sees a different chain here has to be told
where it came from. */}
{!review && claimReviewers?.source === 'task-override' && (
<p className="text-xs text-port-warning">
Seeded<ClaimReviewerSource source={claimReviewers.source} />
</p>
)}
</>
)}
</section>
{submitError && <p role="alert" className="text-xs text-port-error">{submitError}</p>}
<div className="flex items-center justify-end gap-2 pt-2">
<button
type="button"
onClick={onClose}
className="px-3 py-2 bg-port-bg border border-port-border rounded-lg text-sm text-gray-400 hover:text-white transition-colors min-h-[44px]"
>
Cancel
</button>
<button
type="button"
onClick={handleQueue}
disabled={submitting || awaitingPick}
title={awaitingPick ? 'Select a work item first' : undefined}
className="flex items-center gap-1.5 px-3 py-2 bg-blue-500/20 hover:bg-blue-500/30 text-blue-400 border border-blue-500/30 rounded-lg text-sm transition-colors disabled:opacity-50 disabled:cursor-not-allowed min-h-[44px]"
>
{submitting ? <Loader2 size={14} className="animate-spin" /> : <Terminal size={14} />}
Queue {label}
</button>
</div>
</div>
</Drawer>
);
}
export default function SlashDoRunDrawer(props) {
return (
<CodeReviewDefaultsProvider>
<SlashDoRunDrawerBody {...props} />
</CodeReviewDefaultsProvider>
);
}