Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Simplified crystallization strategy for users who have a general research direction (e.g., "I'm interested in LLM reasoning") but lack specificity. Simplifies actor profiling and landscape reconnaissance, then proceeds through direction narrowing, obstacle analysis, goal decomposition, and north-star synthesis. Use when the user's first message reveals a general area but not a specific problem.
.claude/skills/yogsoth-ai-warm-start/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 44% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -20% | 0% |
The user has a general direction — they know the field or area but not the specific problem.
All SOPs in this strategy follow these rules:
| Tactic | Purpose | |--------|---------| | actor-profiling | Understand who the user is | | landscape-reconnaissance | Broad, shallow field exploration | | direction-narrowing | Focus within chosen field(s) | | obstacle-analysis | Identify and mitigate barriers | | goal-decomposition | KAOS-style AND/OR goal structuring | | north-star-synthesis | Converge into North Star + ResearchBrief |
actor-profiling (simplified) → landscape-reconnaissance (simplified or skipped)
→ direction-narrowing → obstacle-analysis → goal-decomposition → north-star-synthesisThis is a reference, not a mandate. How to simplify, how much to simplify, whether to skip entirely — these are your decisions. This strategy suggests simplification as the default posture, but you judge based on what the user's initial message reveals.
You are the general. This strategy gives you:
What you decide:
The only non-negotiable: the process ends with north-star-synthesis producing a North Star + ResearchBrief that the user confirms.
<!-- BEGIN available-tables (generated) -->
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use | | --- | --- | | actor-profiling | Understand who the user is — background, resources, constraints, and deep motivations. Produces an ActorProfile that informs all downstream decisions. Use this tactic at the start of any crystallization process to build a model of the user's capabilities, limitations, and intent. | | direction-narrowing | Focus within the user's chosen field(s). Identify specific sub-directions through deep paper and web research, then present ranked candidates. Use after landscape-reconnaissance has identified fields of interest. | | goal-decomposition | Structure the user's chosen direction into a formal goal tree using KAOS-style AND/OR decomposition. Validate feasibility against ActorProfile and ObstacleReport. Use after obstacle-analysis confirms the direction is viable. | | landscape-reconnaissance | 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. | | north-star-synthesis | Converge all accumulated context into a crystallized North Star statement and structured ResearchBrief. Performs self-review before presenting to user. Use as the final tactic in any start mode — this is where everything comes together. | | obstacle-analysis | 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. |
<!-- 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-01 | fail→fail | 11,081 | 9,696 | -12% | 1 | 1 | 0% | 957 | 1,798 | +88% | 0 | 0 | — |
case-02 | fail→pass | 25,007 | 11,344 | -55% | 1 | 1 | 0% | 2,967 | 2,085 | -30% | 0 | 0 | — |
case-03 | fail→fail | 15,183 | 15,018 | -1% | 1 | 1 | 0% | 2,465 | 2,300 | -7% | 0 | 0 | — |
case-04 | pass→fail | 13,788 | 19,788 | +44% | 1 | 1 | 0% | 2,457 | 3,631 | +48% | 0 | 0 | — |
case-05 | pass→fail | 23,344 | 8,166 | -65% | 1 | 1 | 0% | 2,710 | 2,307 | -15% | 0 | 0 | — |
case-06 | fail→fail | 6,657 | 12,677 | +90% | 1 | 1 | 0% | 211 | 2,369 | +1023% | 0 | 0 | — |
case-07 | fail→fail | 12,972 | 13,042 | +1% | 1 | 1 | 0% | 2,070 | 2,184 | +6% | 0 | 0 | — |
case-08 | fail→pass | 18,075 | 18,283 | +1% | 1 | 1 | 0% | 2,194 | 3,166 | +44% | 0 | 0 | — |
case-09 | fail→pass | 17,017 | 10,638 | -37% | 1 | 1 | 0% | 1,921 | 1,898 | -1% | 0 | 0 | — |
case-10 | fail→pass | 20,278 | 11,937 | -41% | 1 | 1 | 0% | 2,813 | 1,992 | -29% | 0 | 0 | — |
case-11 | fail→pass | 24,201 | 7,209 | -70% | 1 | 1 | 0% | 2,904 | 2,315 | -20% | 0 | 0 | — |
case-12 | pass→pass | 21,833 | 12,042 | -45% | 1 | 1 | 0% | 3,444 | 2,200 | -36% | 0 | 0 | — |
case-13 | pass→pass | 24,223 | 12,887 | -47% | 1 | 1 | 0% | 3,089 | 2,088 | -32% | 0 | 0 | — |
case-14 | pass→pass | 15,870 | 11,238 | -29% | 1 | 1 | 0% | 1,652 | 2,148 | +30% | 0 | 0 | — |
case-15 | pass→pass | 15,641 | 7,031 | -55% | 1 | 1 | 0% | 1,732 | 2,292 | +32% | 0 | 0 | — |
case-16 | pass→pass | 16,264 | 13,544 | -17% | 1 | 1 | 0% | 1,841 | 2,407 | +31% | 0 | 0 | — |
case-17 | fail→fail | 14,237 | 11,607 | -18% | 1 | 1 | 0% | 2,514 | 2,198 | -13% | 0 | 0 | — |
case-18 | fail→pass | 17,103 | 6,877 | -60% | 1 | 1 | 0% | 1,888 | 2,170 | +15% | 0 | 0 | — |
case-19 | fail→fail | 19,344 | 7,476 | -61% | 1 | 1 | 0% | 2,297 | 2,117 | -8% | 0 | 0 | — |
case-20 | pass→fail | 19,578 | 16,995 | -13% | 1 | 1 | 0% | 3,148 | 2,928 | -7% | 0 | 0 | — |
case-21 | fail→fail | 10,456 | 15,938 | +52% | 1 | 1 | 0% | 810 | 2,848 | +252% | 0 | 0 | — |
case-22 | fail→pass | 11,001 | 6,296 | -43% | 1 | 1 | 0% | 1,757 | 1,999 | +14% | 0 | 0 | — |
case-23 | fail→fail | 14,266 | 7,562 | -47% | 1 | 1 | 0% | 1,382 | 2,285 | +65% | 0 | 0 | — |
case-24 | fail→pass | 11,582 | 10,707 | -8% | 1 | 1 | 0% | 902 | 2,018 | +124% | 0 | 0 | — |
case-25 | fail→pass | 19,851 | 11,587 | -42% | 1 | 1 | 0% | 2,236 | 2,227 | -0% | 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. 25 cases were attempted. The headline lift of +24 percentage points is the difference between those two pass rates over the 25 comparable cases. 3 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.