Is my half-built project worth finishing?
Sometimes the half-built thing is a foundation and sometimes it is a hole; the answer comes from a structured read, not loyalty to money already spent. Sunk cost is not a reason to finish; salvageable architecture is.
The longer answer
In practice.
What the read looks at
Whether the core design is sound, what portion genuinely works versus
renders, how documented and testable it is, and whether the missing half is
construction or archaeology. A day or two of reading, a written verdict in
plain words.
The three honest verdicts
Finish it (more exists than you feared); harvest it (keep the design and
data, rebuild the code, cheaper than it sounds with AI leverage); or end it
well (keep the lessons, stop the bleeding). I have delivered all three,
including the third, to people who thanked me later.
What finished means this time
A scope agreed in writing, working software over promises, and evidence at
the end, per rescues. The original dream died
with the original plan; the job is the shortest honest path to something
running.
Same rabbit hole