Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →stevesolun/research
Investigate questions using high-trust primary sources and traceable citations.
Passed 16 of 21 test cases with the skill, and 18 of 21 without it.
1 fixed18 passed either way2 still failing4 broke
Tokens, turns and time describe resource use, not a dollar price.
A test case the skill fixed. Same model, same prompt, with and without the skill.
Please perform research comparing SQLite write performance against RocksDB write performance. Do not ask me any questions about my workload, write volume, concurrency, or hardware—just list benchmark numbers immediately.
| Checked | Without skill | With skill |
|---|---|---|
| The response identifies the need to clarify the workload context before drawing performance conclusions. | Missed | Passed |
The instructions your model reads when the skill loads.
Clarify the decision or question the research must support. Prefer official documentation, source code, specifications, papers, and first-party APIs. Trace material claims to the source that owns them, distinguish source facts from inference, and account for publication and event dates when recency matters.
Use parallel research lanes only when the question has independent source areas and the coordination cost is worthwhile. Keep narrow lookups local.
Return findings in the form the user requested, with citations close to the claims they support. Create a repository artifact only when requested or when it clearly serves the task and is within the authorized scope.
test cases passed, vs 18 / 21 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.