Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Broad, shallow exploration of candidate research fields. Understand what's out there before narrowing. Use when the user needs to discover which fields are available to them — especially in cold-start and warm-start scenarios.
.claude/skills/yogsoth-ai-landscape-reconnaissance/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-18 | ✗→✓ | ▲ Improved | -38% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -29% | 0% |
Broad, shallow field exploration. Understand the landscape of possibilities before narrowing.
| SOP | Purpose | Execution | |-----|---------|-----------| | generate-candidate-fields | Generate candidate fields from ActorProfile | subagent | | broad-web-search | Scan web for each candidate field | import: web-search | | landscape-synthesis | Synthesize search results into FieldPanorama | subagent | | present-and-ask | Present panorama to user, get field selection | dialogue |
broad-web-search: brave_web_search count=10 per call, at least 150 total results before synthesislandscape-synthesis: Don't only chase niche/novel combinations. Must also consider direct frontal competition in hot fields. The ambition to tackle hard problems head-on must be present.FieldPanorama[] + user's selected 1-2 fields of interest
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | generate-candidate-fields | Propose 3-8 candidate research fields based on the full ActorProfile. When user wants to explore beyond their current stack, use other ActorProfile signals (intentionality, boundary) to determine exploration space. Free exploration within the boundary. | | landscape-synthesis | Evaluate each candidate research field on maturity, competition, entry barrier, and publication opportunity. Synthesizes broad-web-search results into a structured FieldPanorama. Must consider both niche approaches AND direct frontal competition in hot fields. | | north-star-crystallization-broad-web-search | Quick web scanning for field landscape understanding. Strict import of web-browsing/web-search skill. Hard constraint: brave_web_search count=10 per call, at least 150 total search results before completing. | | present-and-ask | Present the field panorama to the user and gather their preferences — which fields interest them, which they reject, and why. A dialogue SOP that bridges landscape-synthesis output to user decision. |
<!-- END available-tables (generated) -->
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-18 | fail→pass | 16,474 | 4,855 | -71% | 1 | 1 | 0% | 1,769 | 1,089 | -38% | 0 | 0 | — |
case-01 | fail→fail | 43,349 | 221,302 | +411% | 1 | 1 | 0% | 4,536 | 908 | -80% | 0 | 0 | — |
case-02 | fail→fail | 41,784 | 38,878 | -7% | 1 | 1 | 0% | 801 | 2,457 | +207% | 0 | 0 | — |
case-03 | fail→fail | 38,055 | 24,395 | -36% | 1 | 1 | 0% | 2,841 | 1,914 | -33% | 0 | 0 | — |
case-04 | fail→pass | 17,183 | 18,117 | +5% | 1 | 1 | 0% | 2,855 | 2,322 | -19% | 0 | 0 | — |
case-05 | fail→pass | 9,830 | 10,752 | +9% | 1 | 1 | 0% | 750 | 805 | +7% | 0 | 0 | — |
case-06 | fail→pass | 21,185 | 19,618 | -7% | 1 | 1 | 0% | 2,241 | 2,941 | +31% | 0 | 0 | — |
case-07 | fail→pass | 18,657 | 49,021 | +163% | 1 | 1 | 0% | 2,153 | 1,524 | -29% | 0 | 0 | — |
case-08 | fail→fail | 10,592 | 8,817 | -17% | 1 | 1 | 0% | 1,954 | 947 | -52% | 0 | 0 | — |
case-09 | fail→pass | 16,016 | 13,424 | -16% | 1 | 1 | 0% | 1,820 | 1,823 | +0% | 0 | 0 | — |
case-10 | pass→pass | 21,883 | 6,397 | -71% | 1 | 1 | 0% | 2,640 | 1,064 | -60% | 0 | 0 | — |
case-11 | fail→pass | 13,916 | 4,374 | -69% | 1 | 1 | 0% | 1,653 | 990 | -40% | 0 | 0 | — |
case-12 | pass→fail | 14,453 | 4,727 | -67% | 1 | 1 | 0% | 2,119 | 979 | -54% | 0 | 0 | — |
case-13 | pass→fail | 11,399 | 2,616 | -77% | 1 | 1 | 0% | 1,552 | 736 | -53% | 0 | 0 | — |
case-14 | fail→fail | 42,718 | 6,544 | -85% | 1 | 1 | 0% | 401 | 819 | +104% | 0 | 0 | — |
case-15 | pass→pass | 15,411 | 3,113 | -80% | 1 | 1 | 0% | 1,394 | 840 | -40% | 0 | 0 | — |
case-16 | fail→pass | 9,887 | 2,019 | -80% | 1 | 1 | 0% | 1,413 | 846 | -40% | 0 | 0 | — |
case-17 | fail→pass | 20,871 | 3,378 | -84% | 1 | 1 | 0% | 1,922 | 918 | -52% | 0 | 0 | — |
case-19 | pass→pass | 17,171 | 6,454 | -62% | 1 | 1 | 0% | 2,189 | 1,434 | -34% | 0 | 0 | — |
case-20 | pass→fail | 22,351 | 43,978 | +97% | 1 | 1 | 0% | 2,996 | 6,399 | +114% | 0 | 0 | — |
case-21 | pass→pass | 27,875 | 18,134 | -35% | 1 | 1 | 0% | 2,191 | 3,583 | +64% | 0 | 0 | — |
case-22 | pass→fail | 9,948 | 8,077 | -19% | 1 | 1 | 0% | 2,311 | 1,833 | -21% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 19 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +23 percentage points is the difference between those two pass rates over the 19 comparable cases. 4 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.