Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Identify what blocks the user from pursuing their chosen direction, assess severity, propose mitigations with search-validated evidence, and get user acceptance. Use after direction-narrowing has identified a specific direction.
.claude/skills/yogsoth-ai-obstacle-analysis/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-12 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -43% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -54% | 0% |
Identify barriers, assess severity, propose mitigations, get acceptance.
| SOP | Purpose | Execution | |-----|---------|-----------| | identify-obstacles | Identify obstacles from ActorProfile + chosen direction | subagent (search optional) | | assess-obstacle-severity | Rate severity of each obstacle | subagent (search optional) | | propose-mitigations | Propose evidence-backed mitigations | subagent (search required) | | ask-obstacle-acceptance | Present obstacles + mitigations, get user decision | dialogue (search optional) |
ask-obstacle-acceptance with unresolved obstacles: return to present-candidates (direction-narrowing tactic)ObstacleReport { obstacles[], mitigations[], accepted: bool }
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use | | --- | --- | | ask-obstacle-acceptance | Present obstacles with their severity assessments and proposed mitigations to the user. Ask whether they can accept these obstacles. If unacceptable after 2 rounds, return to present-candidates. | | assess-obstacle-severity | Rate each identified obstacle's difficulty — overcomability, time cost, workaround existence. May optionally use search tools to validate assessments. | | identify-obstacles | Enumerate barriers to pursuing the chosen research direction — knowledge barriers, resource barriers, capability barriers, competition barriers. May optionally use search tools to discover obstacles the user hasn't mentioned. | | propose-mitigations | Propose concrete mitigation strategies for severe obstacles. MUST use search tools to validate that proposed mitigations are realistic — no armchair theorizing. Each mitigation must have evidence of feasibility. |
<!-- 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-12 | fail→pass | 8,732 | 2,436 | -72% | 1 | 1 | 0% | 1,278 | 902 | -29% | 0 | 0 | — |
case-01 | fail→fail | 55,344 | 53,599 | -3% | 1 | 1 | 0% | 8,266 | 1,596 | -81% | 0 | 0 | — |
case-02 | fail→fail | 46,694 | 32,582 | -30% | 1 | 1 | 0% | 6,768 | 1,169 | -83% | 0 | 0 | — |
case-03 | fail→fail | 55,568 | 40,067 | -28% | 1 | 1 | 0% | 8,249 | 1,376 | -83% | 0 | 0 | — |
case-04 | pass→fail | 35,998 | 51,692 | +44% | 1 | 1 | 0% | 3,043 | 5,915 | +94% | 0 | 0 | — |
case-05 | pass→pass | 19,236 | 78,819 | +310% | 1 | 1 | 0% | 1,901 | 8,788 | +362% | 0 | 0 | — |
case-06 | pass→fail | 56,203 | 52,884 | -6% | 1 | 1 | 0% | 5,185 | 3,546 | -32% | 0 | 0 | — |
case-07 | fail→pass | 39,547 | 8,794 | -78% | 1 | 1 | 0% | 1,947 | 1,115 | -43% | 0 | 0 | — |
case-08 | pass→pass | 23,181 | 22,808 | -2% | 1 | 1 | 0% | 2,384 | 1,520 | -36% | 0 | 0 | — |
case-09 | fail→pass | 21,246 | 15,817 | -26% | 1 | 1 | 0% | 2,479 | 2,162 | -13% | 0 | 0 | — |
case-10 | fail→pass | 16,175 | 25,968 | +61% | 1 | 1 | 0% | 2,498 | 3,169 | +27% | 0 | 0 | — |
case-11 | fail→pass | 15,118 | 8,714 | -42% | 1 | 1 | 0% | 2,523 | 1,167 | -54% | 0 | 0 | — |
case-13 | pass→pass | 11,647 | 3,830 | -67% | 1 | 1 | 0% | 1,807 | 1,235 | -32% | 0 | 0 | — |
case-14 | fail→pass | 22,321 | 14,413 | -35% | 1 | 1 | 0% | 2,660 | 2,146 | -19% | 0 | 0 | — |
case-15 | fail→pass | 23,022 | 20,605 | -10% | 1 | 1 | 0% | 2,714 | 2,125 | -22% | 0 | 0 | — |
case-16 | fail→pass | 17,000 | 7,251 | -57% | 1 | 1 | 0% | 1,870 | 908 | -51% | 0 | 0 | — |
case-17 | pass→pass | 22,320 | 9,224 | -59% | 1 | 1 | 0% | 2,456 | 1,144 | -53% | 0 | 0 | — |
case-18 | fail→pass | 16,843 | 8,329 | -51% | 1 | 1 | 0% | 1,474 | 1,114 | -24% | 0 | 0 | — |
case-19 | pass→pass | 12,939 | 7,278 | -44% | 1 | 1 | 0% | 1,195 | 854 | -29% | 0 | 0 | — |
case-20 | fail→pass | 11,680 | 7,969 | -32% | 1 | 1 | 0% | 1,754 | 1,038 | -41% | 0 | 0 | — |
case-21 | fail→pass | 13,981 | 12,278 | -12% | 1 | 1 | 0% | 2,225 | 1,671 | -25% | 0 | 0 | — |
case-22 | fail→pass | 15,343 | 9,754 | -36% | 1 | 1 | 0% | 1,524 | 910 | -40% | 0 | 0 | — |
case-23 | pass→pass | 18,394 | 14,987 | -19% | 1 | 1 | 0% | 2,222 | 2,329 | +5% | 0 | 0 | — |
case-24 | fail→pass | 20,205 | 9,336 | -54% | 1 | 1 | 0% | 2,206 | 1,099 | -50% | 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. 24 cases were attempted, and 20 counted toward the lift figure. The other 4 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 +46 percentage points is the difference between those two pass rates over the 20 comparable cases. 2 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.