Skill Evaluation Engine
§1 Prompt Analysis
Entrada: user_prompt, active_project, available_skills[]
Algoritmo:
- Tokenizar el prompt en keywords
- Para cada skill disponible:
a. Calcular keyword_score = matched_keywords / total_keywords * 100
b. Calcular context_score = project_type_match * 100
c. Calcular history_score = previous_activations_success_rate * 100
d. final_score = keyword_score * 0.4 + context_score * 0.3 + history_score * 0.3
- Filtrar skills con final_score > threshold (default 30)
- Ordenar por final_score descendente
- Retornar top-5
Salida: Lista de skills recomendados con scores y razones
§2 Context Detection
Tipos de proyecto detectables:
- software: presencia de package.json, .sln, Cargo.toml, pom.xml
- research: presencia de experiments/, bibliography/, datasets/
- hardware: presencia de hardware/, bom.json, revisions/
- legal: presencia de legal/, deadlines.json, court-calendar.json
- healthcare: presencia de quality/, pdca/, incidents/
- nonprofit: presencia de impact/, volunteers/
- education: presencia de curricula/, classroom/
Mapping proyecto→skills:
- software → architecture-intelligence, developer-experience
- research → diagram-generation, knowledge-graph
- hardware → regulatory-compliance, cost-management
- legal → cost-management, regulatory-compliance
- healthcare → regulatory-compliance, enterprise-analytics
- nonprofit → executive-reporting, cost-management
§3 Instinct Integration
Cuando un instinto de categoría "context" tiene confianza >70%, boost el score de los skills asociados en +20 puntos.
§4 Feedback Loop
Cada activación registra:
- skill_name, timestamp, prompt_summary, user_accepted (bool)
- Si accepted → +2 al history_score futuro
- Si rejected → -3 al history_score futuro
- Registry:
.opencode/skills/eval-registry.json
1---2name: skill-evaluation3description: Skill Evaluation Engine4---56# Skill Evaluation Engine78## §1 Prompt Analysis910**Entrada**: user_prompt, active_project, available_skills[]1112**Algoritmo**:131. Tokenizar el prompt en keywords142. Para cada skill disponible:15 a. Calcular keyword_score = matched_keywords / total_keywords * 10016 b. Calcular context_score = project_type_match * 10017 c. Calcular history_score = previous_activations_success_rate * 10018 d. final_score = keyword_score * 0.4 + context_score * 0.3 + history_score * 0.3193. Filtrar skills con final_score > threshold (default 30)204. Ordenar por final_score descendente215. Retornar top-52223**Salida**: Lista de skills recomendados con scores y razones2425## §2 Context Detection2627**Tipos de proyecto detectables**:28- software: presencia de package.json, .sln, Cargo.toml, pom.xml29- research: presencia de experiments/, bibliography/, datasets/30- hardware: presencia de hardware/, bom.json, revisions/31- legal: presencia de legal/, deadlines.json, court-calendar.json32- healthcare: presencia de quality/, pdca/, incidents/33- nonprofit: presencia de impact/, volunteers/34- education: presencia de curricula/, classroom/3536**Mapping proyecto→skills**:37- software → architecture-intelligence, developer-experience38- research → diagram-generation, knowledge-graph39- hardware → regulatory-compliance, cost-management40- legal → cost-management, regulatory-compliance41- healthcare → regulatory-compliance, enterprise-analytics42- nonprofit → executive-reporting, cost-management4344## §3 Instinct Integration4546Cuando un instinto de categoría "context" tiene confianza >70%, boost el score de los skills asociados en +20 puntos.4748## §4 Feedback Loop4950Cada activación registra:51- skill_name, timestamp, prompt_summary, user_accepted (bool)52- Si accepted → +2 al history_score futuro53- Si rejected → -3 al history_score futuro54- Registry: `.opencode/skills/eval-registry.json`