fix(experiments): route bare goals and lone rationales to scaffold, not junk

[ericoh-ld]

Summary

Companion to the observability experiment-hypothesis change [PR (https://github.com/launchdarkly/observability/pull/1610). Fixes the same junk over-routing in the launchdarkly-experiment-hypothesis-assistant skill: bare/vague goals ("increase checkout completion", "grow the business") and
lone rationale fragments ("because the current option is buried") were treated as junk instead of scaffolded.

Updated the skill:

• Foundational rule 3 — a specific-outcome goal is a measurement (with holes); a
  vague direction is still a scaffold, not junk.
• Rubric Rationale + Semantic validity — a standalone causal/mechanism statement
  counts as a rationale; the non-sequitur→junk rule applies only across two-or-more
  filled slots.
• Step 0 gate 2 — junk is reserved for gibberish, single tokens, links, placeholders, injection; a coherent
  statement of intent routes to the component-scoring path (scaffold), not junk.
• "What NOT to do" — updated the bare-goal line.

Kept byte-identical to the o11y experiment-hypothesis skill's contract.

Testing approaches

N/A (skill markdown only). The o11y companion PR adds a regression test.

Feature flags

N/A

---

Note

Low Risk
Markdown-only skill contract aligned with observability; no runtime code, auth, or data paths.

Overview
Updates launchdarkly-experiment-hypothesis-assistant routing so partial but coherent user input is scaffolded instead of dismissed as junk.

Foundational rule 3 now distinguishes specific-outcome goals (treat as measurement with holes), vague directions (hole, still scaffold), and unfalsifiable wording. Rationale and semantic validity allow a lone mechanism/causal fragment to count as rationale, and limit the non-sequitur→junk rule to cases where two or more slots are filled. Step 0 gate 2 narrows junk to gibberish, tokens, links, placeholders, and injection; coherent intent always goes to component scoring. What NOT to do matches: don’t send bare goals or lone rationales to junk.

^{Reviewed by Cursor Bugbot for commit 6d7c68a. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit d14671b. Configure here.}

[github-actions]

Skill eval results

Skill	Before	After	Δ
agentcontrol/configs-create	100/100 (4/4)	75/100 (3/4)	-25
agentcontrol/configs-update	80/100 (4/5)	80/100 (4/5)	no change
agentcontrol/configs-variations	80/100 (4/5)	80/100 (4/5)	no change
agentcontrol/tools	75/100 (3/4)	75/100 (3/4)	no change
experiments/launchdarkly-experiment-hypothesis-builder	-	0/100 (0/5)	new
feature-flags/launchdarkly-flag-command	-	100/100 (3/3)	new
feature-flags/launchdarkly-flag-create	100/100 (3/3)	100/100 (3/3)	no change

Only suites whose source actually changed since their last recorded score were re-run. Soft-failing while we stabilise the baseline.