How to Play
A sparse, line-of-sight take on the classic reveal-the-safe-cells puzzle. The hazards are dark nodes you locate and neutralise. Every board — a Vector — is verified solvable by pure deduction: you never have to guess.
The goal
A board is a square grid, up to 12×12. Most cells start hidden — though many Vectors open with a few clues already revealed to get you started. Hidden among them are dark nodes; everything else is safe. Walls are always visible: inert blocks that are never nodes and that stop line-of-sight scans.
Reveal every safe cell and mark every dark node. When both are true, the Vector is SECURED — you win.
The two actions
Tap — reveal
Open a hidden cell. A safe cell shows a hint (a number telling you about nearby nodes), or opens blank.
Tap a dark node and it triggers — you lose the Vector. Read the clues first.
Long-press — mark
Flag a hidden cell you believe is a node. Marking is reversible and never reveals the cell.
Mark the nodes you've deduced, then reveal the rest to secure the Vector.
Good to know
Cascade
Reveal a cell with zero nodes around it and the whole connected blank region opens in one tap — the classic flood.
Marks must be exact
A mark on a safe cell blocks the win until you clear it. The Vector secures only when every real node is marked and nothing else is.
Satisfied hints dim
Once every node a hint counts has been marked, the hint dims on its own — so the clues still worth reading stay bright.
No guessing
Every Vector is solvable by deduction alone. Stuck means a clue is waiting to be read — the Cue tool rings the exact hints that force your next move.
No clock
Nothing is racing you. There's no timer — take the time to read every clue before you tap.
Losing costs coherence
Trigger a node and you lose more than the Vector — it drains your coherence, the sliver of self you jack in with. A clean win costs none, and it regenerates over time, so keep clearing and you can run as long as you like.
Reading the hints
Every revealed clue is one of several kinds. They differ in what they count and which cells they look at. Two scan shapes recur:
Neighbours — the cells directly touching the
clue (8 = all around; 4 = only N/E/S/W).
Line-of-sight beams — straight rays fired outward
to the board edge or the first wall. Dark nodes do not block a beam
— it scans straight through them, counting each one (think x-ray). A
wall stops the beam cold.
Tap a revealed hint to open its description; tap it again to light up the exact cells it scans across the board.
A marked dark node
Wall (blocks line-of-sight beams)
In each example below the relevant nodes are shown marked so you can see them — in real play they stay hidden until you find them.
Hint reference
The toolkit
Surge Guard
Arms a few taps: a tap that would trip a node marks it instead of detonating.
Cue
Rings the revealed hints that force a guaranteed next move. No forced move? It honestly says so and costs nothing.
Threat Map
A quick heat wash estimating each closed cell's node likelihood from what's visible — a heuristic, not a proof.