Fifty-one ideas. Two of them went anywhere.
An idea costs one sentence and no commitment. That's deliberate. It's also why acting on one is rare — and why this page is going to say so instead of finding a friendlier number to lead with.
An idea and a task are not the same shape on purpose.
Saying "what if we tried the cheaper vendor first" costs about as much as saying anything else — one sentence, one tool call, no due date, no owner, no status to keep current. That's create_idea, and it's supposed to be that light. If floating a possibility carried the same weight as committing to it, most possibilities would never get said at all, and the ones that matter would be buried in the ones that got promoted just to justify the effort of writing them down.
A task is the opposite kind of object. It has a due date, which means it shows up unprompted when someone asks what to work on next. An idea has neither. It doesn't expire, doesn't nag, doesn't sort itself into anyone's morning. It is designed to cost nothing to say and nothing to ignore — and it delivers exactly that. Fifty-one of them are sitting in the record right now, uncommitted, exactly as intended, because nothing about the design forces a single one of them into becoming anything else.
What "acted on" actually checks
get_idea_lineage traces whether a specific idea's id shows up as the origin of something real elsewhere in the record — a task, a decision, a shipped changelog entry. An idea counts as acted on because something else points back at it, not because anyone marked it important. That's a small distinction, and it's the one that keeps the number honest.
51
Ideas captured
2
Traceable to something acted on
Two of fifty-one isn't a rate worth turning into a percentage, and this page isn't going to print one. Calibration elsewhere in StratOps refuses to score anything until five decisions have closed, for the same underlying reason: a ratio computed on a handful of cases isn't a measurement, it's noise wearing a decimal point. Read it as two sentences instead of one fraction. Fifty-one ideas were said out loud. Two of them got followed.
The count is narrower than it sounds, and that's the point.
- Not enthusiasmTalking about an idea again, or agreeing it's good, leaves nothing lineage can follow.
- Not a priority flagThere's no star, no upvote, no tag that marks an idea as important. Nothing to set instead of doing the thing.
- Not proximityAn idea mentioned in the same conversation as a task that later shipped isn't linked unless something actually recorded the connection.
- A real link, and only thatA task, a decision, or a changelog entry that exists because the idea did, and can be pointed at.
That's a strict definition on purpose. A looser one would have made 51-to-2 look better without making anything true about the record different. The number stays low because the bar stayed honest, not because the system is bad at connecting things.
This isn't unique to ideas. Observations sit at 190 captured and 0 triaged. Decisions sit at 112 recorded and 0 outcomes ever revisited. Removing the cost of writing something down did not remove the cost of finishing with it — it just moved that cost somewhere nobody's paying it yet.