A deterministic daily draw from the app's own Gurbani corpus that always declines to answer the question you asked, and the reason comes from Gurbani itself.
Once a calendar day, ZindSa draws one passage from its own Gurbani corpus and shows it as today's Vak, for reflection, with its already-vetted translation and its citation.
The draw is deterministic. It is keyed to the day itself, the number of days since a fixed epoch, taken modulo the size of the corpus, so everyone opening the app on the same date sees the same passage, and closing and reopening the app does not reroll it. A shuffle you can retry until you like the answer is a slot machine. This is not that.
It is a personal daily draw from the app's own Guru Granth Sahib corpus. It is not a live feed from a gurdwara. ZindSa does not carry such a feed, and never labels a personal draw as one.
That distinction is easy to blur and profitable to blur, which is why it is written on the public site and not only in a settings screen. An app that showed a passage from its own small offline corpus and captioned it as the day's Hukamnama would be telling its users something false about where the words came from.
The daily reminder is on the same footing: you can switch on a daily notification, and it points you at the reading in the app.
Ask the Sikh mechanism a specific question, about a marriage, a job, a journey, and it does not answer it. Structurally. The module has no path that returns a computed reading; it always returns the refusal object, and the shabad comes attached to the refusal rather than instead of it.
The reason it gives is not a disclaimer written by a lawyer. It is Gurbani's own caution against seeking predictions and calculated auspicious timings, Guru Nanak's critique of pandits calculating muhurats, and the app cites that as the reason the mechanism offers a shabad for reflection rather than a computed reading, for any question asked.
There is a version of this app that computes a Sikh muhurat because the engine to do it already exists two directories away. Not doing that, and saying why, is the whole design.
Ten passages, in Gurmukhi, from four Angs of Guru Granth Sahib: the Mool Mantar and the Slok and first two pauris of Japji Sahib from Ang 1; the closing Slok of Japji Sahib and the opening line of So Dar from Ang 8; the Sukhmani Sahib Mangalacharan from Ang 262; and the first pauri of Anand Sahib from Ang 917.
Every draw comes from that list and each passage carries its own Ang reference. Nothing is generated, paraphrased or reconstructed, the module looks its passage up from the corpus at the moment it draws, so there is no second copy of the text living inside the guidance code to fall out of step.
The gaps are deliberate and recorded: further Japji pauris from the third onward, the rest of Anand Sahib's first five, and the body text of Rehras and Chaupai Sahib were not added in that pass. Not “coming soon”, not added, and written down as not added.
Five gurpurabs ship, and here the app is on less impressive ground than a marketing page would want. Their dates are computed, from the Bikrami lunar tithi, by the same engine that dates every other parva in the app, but they are not computed from a Nanakshahi calculation, because there is no Nanakshahi engine in ZindSa. No sangrand computation exists anywhere in the code.
Two of the five additionally carry a fixed Nanakshahi date, labelled as such, shown beside the Bikrami one. Those two are constants written into the source rather than derived from anything, which is exactly the kind of thing this project tries not to do, and the only reason it survives is that it is disclosed as an alternative reckoning rather than presented as the answer.
The honest summary: the Bikrami and Nanakshahi reckonings genuinely disagree about when some gurpurabs fall, ZindSa computes one of them and shows the other where it has it, and it labels both rather than quietly picking. Each event's provenance note names its own reckoning. What the app does not have is the calculation that would make the Nanakshahi side derivable too.
The Punjabi nakshatra names in the app are still English placeholders, because no Punjabi jantri could be sourced to translate them from. That is a real gap in a Punjabi-language interface and it is not hidden behind a plausible-looking transliteration.
No named scholar and no religious body has reviewed any of this, not the corpus, not the translations, not the refusal logic. The mechanism's restraint comes from reading Gurbani's own words about prediction, which is a reason, not an endorsement.
And the shape of the calendar reflects what could be computed rather than what matters most. Five gurpurabs is not a Sikh year. It is what the tithi engine could date and what could be checked; the rest is not written yet, and no date for writing it is published here, because a delivery promise is not a fact about the app.
Related: what ZindSa shows Sikh users and the daily verse in the FAQ. All posts.