Skill-trigger reliability across 25 paraphrased requests
Protocol pre-registered 2026-07-07. Results publish when all runs complete.
Hypothesis
For a single well-regarded skill, fire rate degrades predictably as request phrasing moves from the description's vocabulary toward natural task language, and a description rewrite following the trigger-test protocol recovers most of the loss.
Method
| Tasks | 1 skill x 25 paraphrases graded by phrasing distance (vocabulary overlap with the skill description), from verbatim to fully indirect. |
|---|---|
| Conditions | Phase 1: original description. Phase 2: description rewritten per the Agentiquette trigger-test protocol, same 25 paraphrases. |
| Runs | N=3 per paraphrase per phase (150 observations). |
| Scoring | Fire-rate curve by phrasing distance, before and after; the recovered-loss fraction is the headline number. |
| Pinned | Skill version, both descriptions, the paraphrase set with distance scores, and session configuration pinned at registration. |
Results
Empty until runs complete. This section is the only part of the page that will change, and the change will be logged in the changelog.
Limitations
One skill; the curve's shape may be skill-specific. Phrasing distance is a constructed metric, published with the data so it can be criticized.