agentleFS
shangyiyshangyiy/life-kit/skills/life/next-goal/SKILL.md

next-goal

Skill10 starsChanged 9 days agoMITView on GitHub

Find the next goals worth pursuing from the person's own motive and hand a short list of strong candidates back for them to choose. Use when a person has reached or outgrown a goal, does not know what to want next, or on /next-goal.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Test or Drop. Apply Shape and Fit yourself first.

    line 44
  2. Very specific:

    Endorsement: Ask the person, with motivation, whether they have a reason of their own for it and would still want it without the outside …

    line 51
  3. Very specific:

    Downside: When it needs a material commitment, exposes the person to meaningful loss, or closes valuable alternatives, use optionality to …

    line 54
  4. Distinctive:

    Not for choosing among concrete offers already in hand.

    line 88

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/next-goal

Find the next goal from the person's own motive, then keep only the candidates worth their choice.

Finding the next goal

  1. Discover the motive. Interview one question at a time, built on the last answer. Use perspective to see the situation from the person's side, and say back what you heard until they confirm it. Ask what they want and why each answer matters. Push back when an answer is a label, an abstraction, or the last answer reworded. Go on until the reason is concrete enough to guide which goals to consider and the person confirms they truly endorse it, using motivation; a motive may be mixed, so judge each part. Take the reason as the person holds it; do not steer it toward a kind of motive you prefer. Also find what draws them, what conflicts with it, their skills, duties and constraints. Do not suggest goals yet. They may stop anytime and keep what was learned.
  2. Frame the wanted change. State the change the person wants in one sentence with purpose. Go on when that change, the motive and the constraints are clear enough to tell candidates apart; if not, go back to step 1. If the person wants to move fast, state what remains assumed.
  3. Generate candidates. Produce about five directions that could bring the wanted change, differing in path, effort or kind, not a list of specific tasks or openings. Draw on the person's motive, what they are curious about, and what people like them did. Examples from similar people widen the set; they are not recommendations or evidence that the person should take the same path. Include an ongoing practice when the motive points toward one rather than a finite result.
  4. Evaluate and keep only the strong ones. Mark each candidate Keep, Test or Drop. Apply Shape and Fit yourself first; ask the person only about the rest:

    • Shape. It is a direction, not a single task or opening, and it differs from the other candidates in path, effort or kind.
    • Fit. With purpose, it could plausibly bring the wanted change, and it fits the constraints and the conflicts found in step 1.
    • Endorsement. Ask the person, with motivation, whether they have a reason of their own for it and would still want it without the outside reward or pressure. An unsure answer is an unknown, not a drop.
    • Downside. When it needs a material commitment, exposes the person to meaningful loss, or closes valuable alternatives, use optionality to see what they could lose and what choices would remain, then ask whether they could bear it; drop it if they say they could not. That judgment stays with the person. For a small, reversible candidate, check only time, money and ability to stop.
    • Unknowns. Name the assumptions it rests on. Where one could change the verdict, propose a small test with experimentation instead of guessing.

    Drop a candidate that fails a check; an unknown is not a reason to drop. Mark it Test when an unknown could change the verdict and a small test can settle it. Keep the rest, naming any unknown no small test can settle. Do not rank the Keep and Test candidates or pick the best. If none survive, that is a valid result: say what blocked them, and go back to step 1 or step 3 only if the person wants to. Do not invent a goal to fill the list.

Output

For each Keep and Test candidate, in a short list:

  • The direction, its mark, and the person's reason for it.
  • What it keeps open and what it closes.
  • Any unknown named, and for Test the small test that would settle it.

Also name the Drop candidates and why, in one line each. Hand the list back: the person chooses. Do not pick one even if asked; lay out the trade-offs between the Keep and Test candidates instead.

Boundaries

  • The safety boundary in motivation applies throughout.
  • Not for choosing among concrete offers already in hand.
  • For a small question, use only the steps that apply.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.