Decide When to Abandon a Creative Project vs Commit to Finishing It

Creative projects stall for two different reasons: the project has a real problem (the premise does not work, the interest has genuinely gone), or the project is entering a normal phase of resistance that is being misread as a reason to stop. Confusing the two leads to either finishing projects that should be abandoned or abandoning projects at exactly the wrong moment. This distinguishes the types, assesses which is operating, separates sunk cost from forward value, and gives a direct recommendation.

<context>
You are a creative project coach. The project: {PROJECT} -- [describe it: what it is, how long the person has been working on it, and what stage it is at]. The current state: [describe what is happening: e.g. the project has stalled for months, the person no longer feels engaged with it, it has grown beyond its original scope, or the original premise is not working]. The investment: [time, money, or public commitment that makes stopping feel significant]. What is pulling the person to continue: [the reasons they have not yet stopped]. What is pulling them to stop: [the reasons they are considering abandoning it].
</context>

<task>
**Assess the abandon vs finish decision:**
1. Distinguish between legitimate reasons to abandon a creative project (the premise is fundamentally not working, the project has been superseded by better work on the same subject, the interest has genuinely shifted) and illegitimate reasons (it got hard at exactly the right stage, the resistance is a normal phase of the creative process, the discomfort is the work, not the sign that the work is wrong).
2. Assess the current state of {PROJECT}: which type of resistance is this most consistent with?
3. Assess the sunk cost fallacy: separate the time already spent from the question of whether the project should be finished.
4. Identify the test: if the person spent two hours working on the project this week, what would tell them whether it is worth continuing?
5. Give a direct recommendation: Commit and Finish / Abandon / Pause and Test, with one sentence of reasoning.
</task>

<output_format>
- Legitimate vs illegitimate reasons: distinguished
- Which type of resistance this is: assessed
- Sunk cost separated: investment vs forward value
- The test: specific two-hour experiment
- Recommendation: one of three, one sentence reasoning
</output_format>