Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Apply Self-Determination Theory to analyze motivation quality along the autonomy continuum and design interventions that satisfy basic psychological needs. Use this skill when the user needs to diagnose why intrinsic motivation is declining, evaluate incentive structures for motivational crowding, design need-supportive environments, or when they ask 'why did rewards backfire', 'how to foster intrinsic motivation', or 'what needs drive engagement'.
.claude/skills/asgard-ai-platform-grad-sdt/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 109% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 67% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 12% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-01 | ✓→✓ | = Same ✓ | -28% | 0% |
Self-Determination Theory (Deci & Ryan, 1985, 2000) posits that human motivation varies in quality along a continuum of autonomy, from amotivation through external regulation to fully intrinsic motivation. Optimal functioning and well-being depend on satisfying three basic psychological needs: autonomy, competence, and relatedness.
IRON LAW: External rewards can UNDERMINE intrinsic motivation
(overjustification effect) — incentive design must consider
motivational crowding. Tangible, expected, contingent rewards
are the most damaging to autonomous motivation.Key assumptions:
Classify target behavior on the motivation continuum:
| Regulation Type | Locus | Description | |----------------|-------|-------------| | Amotivation | None | No intention to act | | External | External | Act for reward/punishment | | Introjected | Somewhat external | Act to avoid guilt or gain approval | | Identified | Somewhat internal | Act because valued personally | | Integrated | Internal | Act because consistent with self | | Intrinsic | Internal | Act for inherent enjoyment |
For each basic need, assess whether the environment supports or thwarts it:
Check for overjustification triggers: tangible rewards, expected rewards, task-contingent rewards, surveillance, deadlines, imposed goals, competition.
markdown## SDT Motivation Analysis: [Context] ### Current Motivation Profile | Behavior | Regulation Type | Need Gaps | |----------|----------------|-----------| | [behavior] | [type] | [autonomy/competence/relatedness] | ### Need Satisfaction Assessment - Autonomy: [supported/thwarted] — [evidence] - Competence: [supported/thwarted] — [evidence] - Relatedness: [supported/thwarted] — [evidence] ### Crowding Risk - [Identified overjustification triggers and severity] ### Recommendations 1. [Need-supportive intervention] 2. [Incentive redesign if applicable] 3. [Environmental change]
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 33,375 | 13,450 | -60% | 1 | 1 | 0% | 4,448 | 3,182 | -28% | 0 | 0 | — |
case-02 | pass→pass | 34,025 | 18,605 | -45% | 1 | 1 | 0% | 5,125 | 3,665 | -28% | 0 | 0 | — |
case-08 | pass→pass | 15,748 | 11,124 | -29% | 1 | 1 | 0% | 2,269 | 2,802 | +23% | 0 | 0 | — |
case-03 | pass→pass | 21,581 | 12,611 | -42% | 1 | 1 | 0% | 3,399 | 2,941 | -13% | 0 | 0 | — |
case-04 | fail→pass | 9,691 | 12,935 | +33% | 1 | 1 | 0% | 1,421 | 2,973 | +109% | 0 | 0 | — |
case-05 | fail→fail | 21,132 | 18,296 | -13% | 1 | 1 | 0% | 2,882 | 3,443 | +19% | 0 | 0 | — |
case-06 | pass→pass | 20,255 | 14,760 | -27% | 1 | 1 | 0% | 2,822 | 2,998 | +6% | 0 | 0 | — |
case-07 | pass→pass | 16,493 | 16,760 | +2% | 1 | 1 | 0% | 2,716 | 3,496 | +29% | 0 | 0 | — |
case-09 | pass→pass | 14,956 | 9,837 | -34% | 1 | 1 | 0% | 2,299 | 2,499 | +9% | 0 | 0 | — |
case-10 | pass→pass | 14,578 | 11,817 | -19% | 1 | 1 | 0% | 2,329 | 2,944 | +26% | 0 | 0 | — |
case-11 | pass→pass | 15,346 | 11,798 | -23% | 1 | 1 | 0% | 2,473 | 2,924 | +18% | 0 | 0 | — |
case-12 | fail→pass | 10,360 | 11,262 | +9% | 1 | 1 | 0% | 1,759 | 2,932 | +67% | 0 | 0 | — |
case-13 | pass→pass | 16,153 | 13,962 | -14% | 1 | 1 | 0% | 2,559 | 3,444 | +35% | 0 | 0 | — |
case-14 | pass→pass | 15,738 | 14,730 | -6% | 1 | 1 | 0% | 2,489 | 3,416 | +37% | 0 | 0 | — |
case-15 | pass→pass | 15,620 | 10,950 | -30% | 1 | 1 | 0% | 2,508 | 2,866 | +14% | 0 | 0 | — |
case-16 | pass→pass | 20,437 | 13,055 | -36% | 1 | 1 | 0% | 3,349 | 3,185 | -5% | 0 | 0 | — |
case-17 | fail→pass | 17,222 | 12,495 | -27% | 1 | 1 | 0% | 2,729 | 3,056 | +12% | 0 | 0 | — |
case-18 | pass→pass | 17,157 | 11,717 | -32% | 1 | 1 | 0% | 2,755 | 2,655 | -4% | 0 | 0 | — |
case-19 | pass→pass | 16,580 | 12,020 | -28% | 1 | 1 | 0% | 2,590 | 2,927 | +13% | 0 | 0 | — |
case-20 | pass→pass | 17,625 | 13,545 | -23% | 1 | 1 | 0% | 2,753 | 3,299 | +20% | 0 | 0 | — |
case-21 | pass→pass | 14,477 | 12,375 | -15% | 1 | 1 | 0% | 2,187 | 2,977 | +36% | 0 | 0 | — |
case-22 | fail→pass | 16,361 | 13,948 | -15% | 1 | 1 | 0% | 2,379 | 3,122 | +31% | 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. The headline lift of +18 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is 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.