agentleFS
agiea-brainagiea-brain/agency-skills/skills/tapajyoti-das/quarter-goal-filter/SKILL.md

quarter-goal-filter

Skill1 starsChanged 17 days agoMITView on GitHub

Use this skill to lock one goal for the quarter, then to judge any new idea, channel, pitch or partnership against it before it eats a month. Two modes: setup produces the goal in one sentence with a measure and an explicit not-doing list; filter produces a one-line verdict on an idea, what it displaces, and a sentence the owner can say out loud. Also use for "should we even be doing this", "set the quarterly goal", "is this a distraction", "what are we saying no to".

The file

Show its rules and text

We split this skill into the rules it gives an agent

17 rules. About 0 are common, 17 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈17

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Check what was committed last quarter and what happened before setting anything new.

    line 22
  2. Distinctive:

    Write the measure separately, so there is no argument in April about whether it happened.

    line 26
  3. Distinctive:

    Read the goal and check whether this idea, or something close to it, has been tried before.

    line 34
  4. Distinctive:

    If it serves the goal: say what it needs and what it displaces.

    line 38
  5. Distinctive:

    If it is neutral: say so plainly and ask whether there is a reason to do it anyway.

    line 40
  6. Distinctive:

    If it competes: say what it costs the goal, and name the version of it that would be worth doing next quarter instead.

    line 42
  7. Distinctive:

    MUST write the goal with a number and a date, so it can be failed.

    line 56
  8. Distinctive:

    MUST produce an explicit not-doing list at setup.

    line 57
  9. Distinctive:

    MUST name one binding constraint, not several.

    line 58
  10. Distinctive:

    MUST give exactly one of three verdicts on a filter run.

    line 59
  11. Distinctive:

    MUST state what an approved idea displaces.

    line 60
  12. Distinctive:

    MUST surface prior attempts at the same idea.

    line 61
  13. Distinctive:

    MUST give a repeatable one-line answer.

    line 62
  14. Distinctive:

    NEVER soften a verdict to be agreeable.

    line 63
  15. Distinctive:

    NEVER approve something because it is good in general.

    line 64
  16. Distinctive:

    NEVER let the goal be rewritten mid-filter to fit an idea.

    line 65
  17. Distinctive:

    NEVER produce a score. A number invites negotiation with itself.

    line 66

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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

Applies to a small firm losing quarters to individually reasonable decisions. Run setup once a quarter, filter whenever something new arrives.

The play

Setup

  1. Check what was committed last quarter and what happened before setting anything new. A firm that has set and missed the same goal three times has a capacity problem, not a goal problem, and setting it a fourth time will not fix it.
  2. Get the goal to one sentence with a number and a date in it. "Grow revenue" is not a goal. "Three new retainers at or above the floor by March 31" is, because it can be failed.
  3. Write the measure separately, so there is no argument in April about whether it happened.
  4. Write what the firm is explicitly not doing this quarter. A goal with nothing given up for it is a wish, and it will not survive the first interesting inbound. This list is the part that does the work, and it is the part people skip.
  5. Name the constraint that will actually bind: a specific person's time, cash, or a client dependency. Not all three. The one that breaks first.

Filter

  1. Read the goal and check whether this idea, or something close to it, has been tried before. If it has, lead with what happened. Firms re-propose their own abandoned experiments constantly, and nobody remembers.
  2. Give one of three verdicts and do not hedge into a fourth: it serves the goal, it is neutral, or it competes with the goal.
  3. If it serves the goal: say what it needs and what it displaces. Everything displaces something, and the displacement is the real cost.
  4. If it is neutral: say so plainly and ask whether there is a reason to do it anyway. Some neutral things are worth doing. Most are what quarters disappear into, one reasonable yes at a time.
  5. If it competes: say what it costs the goal, and name the version of it that would be worth doing next quarter instead. Killing an idea outright invites the argument again next week. Scheduling it ends it.
  6. Give the owner a one-line answer they can repeat to whoever brought the idea, without having to reconstruct the reasoning.

What good looks like

  • First thing I check that most people skip: the not-doing list from last quarter, and whether we did those things anyway. That list is the only honest record of whether the goal ever held.
  • The mediocre version: a goal with four parts, which is four goals and therefore none. Everything qualifies as progress and the quarter ends with motion and no result.
  • I know it is right when I can turn down something genuinely good in one sentence, without reopening the whole strategy.
  • Second test: the goal can be failed. If there is no outcome that counts as missing it, it is a direction, not a goal.
  • Red flag: the goal gets edited mid-filter to accommodate whatever just arrived. That is not a decision, it is a rationalisation with a calendar on it.

Rules

  • MUST write the goal with a number and a date, so it can be failed.
  • MUST produce an explicit not-doing list at setup.
  • MUST name one binding constraint, not several.
  • MUST give exactly one of three verdicts on a filter run.
  • MUST state what an approved idea displaces.
  • MUST surface prior attempts at the same idea.
  • MUST give a repeatable one-line answer.
  • NEVER soften a verdict to be agreeable.
  • NEVER approve something because it is good in general.
  • NEVER let the goal be rewritten mid-filter to fit an idea. Changing the goal is its own decision.
  • NEVER produce a score. A number invites negotiation with itself.

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.