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
/**
* Season/episode structure recommendation.
*
* Comic-as-TV norms: 6–10 issues per season/volume is the sweet spot. Single
* volume up to 12 (the classic 12-issue maxi-series); 3-volume arc lands
* roughly at 18–30 issues; 4+ volumes only past 30. The "3 × 8 = 24" point
* is the canonical "3-season arc" target.
*
* Used both as renderer input to the arc-overview LLM prompt (so the LLM
* stops slicing 12-issue runs into three 4-issue seasons) and as the
* client-side hint shown next to the issue-count input.
*/
export const SEASON_MIN_EPISODES = 6;
export const SEASON_MAX_EPISODES = 11;
/**
* Pick the natural season count for a given total. Boundaries are tuned so
* 12 → 1 season, 18 → 3, 24 → 3, 30 → 3, 36 → 4.
*/
function pickSeasonCount(total) {
if (total <= 12) return 1;
if (total <= 17) return 2;
if (total <= 32) return 3;
if (total <= 44) return 4;
return 5;
}
/**
* Compute the recommended structure for a total issue/episode count.
*
* recommendStructure(12) → { seasons: 1, perSeason: [12] }
* recommendStructure(24) → { seasons: 3, perSeason: [8, 8, 8] }
* recommendStructure(20) → { seasons: 3, perSeason: [7, 7, 6] }
*
* Returns `null` for zero/negative input so callers can short-circuit
* (e.g. the UI hint hides itself when the user hasn't entered a target yet).
*/
export function recommendStructure(total) {
const n = Math.floor(Number(total) || 0);
if (n <= 0) return null;
const seasons = pickSeasonCount(n);
const base = Math.floor(n / seasons);
const remainder = n % seasons;
// Front-load extras into the earlier seasons so the array stays
// monotonically non-increasing — a 22-issue run reads as [8,7,7] not [7,7,8].
const perSeason = Array.from(
{ length: seasons },
(_, i) => base + (i < remainder ? 1 : 0),
);
return { seasons, perSeason };
}
/**
* Human-readable summary like "3 volumes × 8 episodes" or "2 volumes
* × ~9 (9, 9)". Used in the LLM prompt and the client hint.
*/
export function describeStructure(structure) {
if (!structure) return '';
const { seasons, perSeason } = structure;
const allSame = perSeason.every((n) => n === perSeason[0]);
if (allSame) {
return seasons === 1
? `1 volume × ${perSeason[0]} episodes`
: `${seasons} volumes × ${perSeason[0]} episodes`;
}
return `${seasons} volumes × ~${Math.round(perSeason.reduce((a, b) => a + b, 0) / seasons)} (${perSeason.join(', ')})`;
}