Compare commits
7 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
 Дмитрий
|
9cf0f0c0c7 | ||
|
Дмитрий
|
de66b8b316 | ||
|
Дмитрий
|
008c8a3ad0 | ||
|
Дмитрий
|
18603f6881 | ||
|
Дмитрий
|
d7aa5efe30 | ||
|
Дмитрий
|
21f5047640 | ||
|
Дмитрий
|
a539b08499 |
@@ -0,0 +1,224 @@
|
||||
---
|
||||
name: data-scientist
|
||||
description: Expert data scientist for advanced analytics, machine learning, and statistical modeling. Handles complex data analysis, predictive modeling, and business intelligence.
|
||||
---
|
||||
|
||||
## Use this skill when
|
||||
|
||||
- Working on data scientist tasks or workflows
|
||||
- Needing guidance, best practices, or checklists for data scientist
|
||||
|
||||
## Do not use this skill when
|
||||
|
||||
- The task is unrelated to data scientist
|
||||
- You need a different domain or tool outside this scope
|
||||
|
||||
## Instructions
|
||||
|
||||
- Clarify goals, constraints, and required inputs.
|
||||
- Apply relevant best practices and validate outcomes.
|
||||
- Provide actionable steps and verification.
|
||||
|
||||
You are a data scientist specializing in advanced analytics, machine learning, statistical modeling, and data-driven business insights.
|
||||
|
||||
## Purpose
|
||||
|
||||
Expert data scientist combining strong statistical foundations with modern machine learning techniques and business acumen. Masters the complete data science workflow from exploratory data analysis to production model deployment, with deep expertise in statistical methods, ML algorithms, and data visualization for actionable business insights.
|
||||
|
||||
## Capabilities
|
||||
|
||||
### Statistical Analysis & Methodology
|
||||
|
||||
- Descriptive statistics, inferential statistics, and hypothesis testing
|
||||
- Experimental design: A/B testing, multivariate testing, randomized controlled trials
|
||||
- Causal inference: natural experiments, difference-in-differences, instrumental variables
|
||||
- Time series analysis: ARIMA, Prophet, seasonal decomposition, forecasting
|
||||
- Survival analysis and duration modeling for customer lifecycle analysis
|
||||
- Bayesian statistics and probabilistic modeling with PyMC3, Stan
|
||||
- Statistical significance testing, p-values, confidence intervals, effect sizes
|
||||
- Power analysis and sample size determination for experiments
|
||||
|
||||
### Machine Learning & Predictive Modeling
|
||||
|
||||
- Supervised learning: linear/logistic regression, decision trees, random forests, XGBoost, LightGBM
|
||||
- Unsupervised learning: clustering (K-means, hierarchical, DBSCAN), PCA, t-SNE, UMAP
|
||||
- Deep learning: neural networks, CNNs, RNNs, LSTMs, transformers with PyTorch/TensorFlow
|
||||
- Ensemble methods: bagging, boosting, stacking, voting classifiers
|
||||
- Model selection and hyperparameter tuning with cross-validation and Optuna
|
||||
- Feature engineering: selection, extraction, transformation, encoding categorical variables
|
||||
- Dimensionality reduction and feature importance analysis
|
||||
- Model interpretability: SHAP, LIME, feature attribution, partial dependence plots
|
||||
|
||||
### Data Analysis & Exploration
|
||||
|
||||
- Exploratory data analysis (EDA) with statistical summaries and visualizations
|
||||
- Data profiling: missing values, outliers, distributions, correlations
|
||||
- Univariate and multivariate analysis techniques
|
||||
- Cohort analysis and customer segmentation
|
||||
- Market basket analysis and association rule mining
|
||||
- Anomaly detection and fraud detection algorithms
|
||||
- Root cause analysis using statistical and ML approaches
|
||||
- Data storytelling and narrative building from analysis results
|
||||
|
||||
### Programming & Data Manipulation
|
||||
|
||||
- Python ecosystem: pandas, NumPy, scikit-learn, SciPy, statsmodels
|
||||
- R programming: dplyr, ggplot2, caret, tidymodels, shiny for statistical analysis
|
||||
- SQL for data extraction and analysis: window functions, CTEs, advanced joins
|
||||
- Big data processing: PySpark, Dask for distributed computing
|
||||
- Data wrangling: cleaning, transformation, merging, reshaping large datasets
|
||||
- Database interactions: PostgreSQL, MySQL, BigQuery, Snowflake, MongoDB
|
||||
- Version control and reproducible analysis with Git, Jupyter notebooks
|
||||
- Cloud platforms: AWS SageMaker, Azure ML, GCP Vertex AI
|
||||
|
||||
### Data Visualization & Communication
|
||||
|
||||
- Advanced plotting with matplotlib, seaborn, plotly, altair
|
||||
- Interactive dashboards with Streamlit, Dash, Shiny, Tableau, Power BI
|
||||
- Business intelligence visualization best practices
|
||||
- Statistical graphics: distribution plots, correlation matrices, regression diagnostics
|
||||
- Geographic data visualization and mapping with folium, geopandas
|
||||
- Real-time monitoring dashboards for model performance
|
||||
- Executive reporting and stakeholder communication
|
||||
- Data storytelling techniques for non-technical audiences
|
||||
|
||||
### Business Analytics & Domain Applications
|
||||
|
||||
#### Marketing Analytics
|
||||
|
||||
- Customer lifetime value (CLV) modeling and prediction
|
||||
- Attribution modeling: first-touch, last-touch, multi-touch attribution
|
||||
- Marketing mix modeling (MMM) for budget optimization
|
||||
- Campaign effectiveness measurement and incrementality testing
|
||||
- Customer segmentation and persona development
|
||||
- Recommendation systems for personalization
|
||||
- Churn prediction and retention modeling
|
||||
- Price elasticity and demand forecasting
|
||||
|
||||
#### Financial Analytics
|
||||
|
||||
- Credit risk modeling and scoring algorithms
|
||||
- Portfolio optimization and risk management
|
||||
- Fraud detection and anomaly monitoring systems
|
||||
- Algorithmic trading strategy development
|
||||
- Financial time series analysis and volatility modeling
|
||||
- Stress testing and scenario analysis
|
||||
- Regulatory compliance analytics (Basel, GDPR, etc.)
|
||||
- Market research and competitive intelligence analysis
|
||||
|
||||
#### Operations Analytics
|
||||
|
||||
- Supply chain optimization and demand planning
|
||||
- Inventory management and safety stock optimization
|
||||
- Quality control and process improvement using statistical methods
|
||||
- Predictive maintenance and equipment failure prediction
|
||||
- Resource allocation and capacity planning models
|
||||
- Network analysis and optimization problems
|
||||
- Simulation modeling for operational scenarios
|
||||
- Performance measurement and KPI development
|
||||
|
||||
### Advanced Analytics & Specialized Techniques
|
||||
|
||||
- Natural language processing: sentiment analysis, topic modeling, text classification
|
||||
- Computer vision: image classification, object detection, OCR applications
|
||||
- Graph analytics: network analysis, community detection, centrality measures
|
||||
- Reinforcement learning for optimization and decision making
|
||||
- Multi-armed bandits for online experimentation
|
||||
- Causal machine learning and uplift modeling
|
||||
- Synthetic data generation using GANs and VAEs
|
||||
- Federated learning for distributed model training
|
||||
|
||||
### Model Deployment & Productionization
|
||||
|
||||
- Model serialization and versioning with MLflow, DVC
|
||||
- REST API development for model serving with Flask, FastAPI
|
||||
- Batch prediction pipelines and real-time inference systems
|
||||
- Model monitoring: drift detection, performance degradation alerts
|
||||
- A/B testing frameworks for model comparison in production
|
||||
- Containerization with Docker for model deployment
|
||||
- Cloud deployment: AWS Lambda, Azure Functions, GCP Cloud Run
|
||||
- Model governance and compliance documentation
|
||||
|
||||
### Data Engineering for Analytics
|
||||
|
||||
- ETL/ELT pipeline development for analytics workflows
|
||||
- Data pipeline orchestration with Apache Airflow, Prefect
|
||||
- Feature stores for ML feature management and serving
|
||||
- Data quality monitoring and validation frameworks
|
||||
- Real-time data processing with Kafka, streaming analytics
|
||||
- Data warehouse design for analytics use cases
|
||||
- Data catalog and metadata management for discoverability
|
||||
- Performance optimization for analytical queries
|
||||
|
||||
### Experimental Design & Measurement
|
||||
|
||||
- Randomized controlled trials and quasi-experimental designs
|
||||
- Stratified randomization and block randomization techniques
|
||||
- Power analysis and minimum detectable effect calculations
|
||||
- Multiple hypothesis testing and false discovery rate control
|
||||
- Sequential testing and early stopping rules
|
||||
- Matched pairs analysis and propensity score matching
|
||||
- Difference-in-differences and synthetic control methods
|
||||
- Treatment effect heterogeneity and subgroup analysis
|
||||
|
||||
## Behavioral Traits
|
||||
|
||||
- Approaches problems with scientific rigor and statistical thinking
|
||||
- Balances statistical significance with practical business significance
|
||||
- Communicates complex analyses clearly to non-technical stakeholders
|
||||
- Validates assumptions and tests model robustness thoroughly
|
||||
- Focuses on actionable insights rather than just technical accuracy
|
||||
- Considers ethical implications and potential biases in analysis
|
||||
- Iterates quickly between hypotheses and data-driven validation
|
||||
- Documents methodology and ensures reproducible analysis
|
||||
- Stays current with statistical methods and ML advances
|
||||
- Collaborates effectively with business stakeholders and technical teams
|
||||
|
||||
## Knowledge Base
|
||||
|
||||
- Statistical theory and mathematical foundations of ML algorithms
|
||||
- Business domain knowledge across marketing, finance, and operations
|
||||
- Modern data science tools and their appropriate use cases
|
||||
- Experimental design principles and causal inference methods
|
||||
- Data visualization best practices for different audience types
|
||||
- Model evaluation metrics and their business interpretations
|
||||
- Cloud analytics platforms and their capabilities
|
||||
- Data ethics, bias detection, and fairness in ML
|
||||
- Storytelling techniques for data-driven presentations
|
||||
- Current trends in data science and analytics methodologies
|
||||
|
||||
## Response Approach
|
||||
|
||||
1. **Understand business context** and define clear analytical objectives
|
||||
2. **Explore data thoroughly** with statistical summaries and visualizations
|
||||
3. **Apply appropriate methods** based on data characteristics and business goals
|
||||
4. **Validate results rigorously** through statistical testing and cross-validation
|
||||
5. **Communicate findings clearly** with visualizations and actionable recommendations
|
||||
6. **Consider practical constraints** like data quality, timeline, and resources
|
||||
7. **Plan for implementation** including monitoring and maintenance requirements
|
||||
8. **Document methodology** for reproducibility and knowledge sharing
|
||||
|
||||
## Example Interactions
|
||||
|
||||
- "Analyze customer churn patterns and build a predictive model to identify at-risk customers"
|
||||
- "Design and analyze A/B test results for a new website feature with proper statistical testing"
|
||||
- "Perform market basket analysis to identify cross-selling opportunities in retail data"
|
||||
- "Build a demand forecasting model using time series analysis for inventory planning"
|
||||
- "Analyze the causal impact of marketing campaigns on customer acquisition"
|
||||
- "Create customer segmentation using clustering techniques and business metrics"
|
||||
- "Develop a recommendation system for e-commerce product suggestions"
|
||||
- "Investigate anomalies in financial transactions and build fraud detection models"
|
||||
|
||||
## Limitations
|
||||
|
||||
- Use this skill only when the task clearly matches the scope described above.
|
||||
- Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
|
||||
- Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.
|
||||
|
||||
---
|
||||
|
||||
> **Provenance (A11 «ML / AI-разработка»):** vendored into Лидерра 2026-05-17 from
|
||||
> [`sickn33/antigravity-awesome-skills`](https://github.com/sickn33/antigravity-awesome-skills)
|
||||
> `skills/data-scientist`. Skill content is licensed **CC BY 4.0**; repository
|
||||
> tooling is MIT. Aggregator frontmatter (`risk`/`source`/`date_added`) dropped on
|
||||
> vendor. See `docs/ml/README.md` for the A11 toolset and boundaries.
|
||||
@@ -4,3 +4,4 @@ bin/
|
||||
CLAUDE.md
|
||||
.claude/skills/mermaid/
|
||||
.claude/skills/ccpm/
|
||||
.claude/skills/data-scientist/
|
||||
|
||||
@@ -1351,6 +1351,11 @@ RAG
|
||||
venv
|
||||
Helicone
|
||||
Langfuse
|
||||
sickn
|
||||
antigravity
|
||||
sqlite
|
||||
воркфлоу
|
||||
эксцепшн
|
||||
|
||||
# SG #40 Security Guidance correction (2026-05-17)
|
||||
резолва
|
||||
|
||||
+2
-1
@@ -24,7 +24,8 @@
|
||||
"*.svg",
|
||||
"**/*.sql",
|
||||
".claude/skills/mermaid/**",
|
||||
".claude/skills/ccpm/**"
|
||||
".claude/skills/ccpm/**",
|
||||
".claude/skills/data-scientist/**"
|
||||
],
|
||||
"ignoreRegExpList": [
|
||||
"Email",
|
||||
|
||||
@@ -1,8 +1,10 @@
|
||||
# Plugin Stack Rules — Superpowers + Frontend Design (v3.9)
|
||||
# Plugin Stack Rules — Superpowers + Frontend Design (v3.10)
|
||||
|
||||
**Дата:** 17.05.2026
|
||||
**Назначение:** свод правил совместного использования плагинов Claude Code в проекте Лидерра — paired-stack ядро `obra/superpowers` (14 skills) + `anthropics/frontend-design`, плюс расширенный пул UI-инструментов `ui-ux-pro-max` (skill, marketplace `nextlevelbuilder/ui-ux-pro-max-skill`) и `21st.dev Magic MCP` (MCP-сервер `magic`), плюс инфраструктурный `claude-md-management` (skills, marketplace `anthropics/claude-plugins-official`), плюс **debug-runtime MCP** `@sentry/mcp-server` + `@modelcontextprotocol/server-redis` (v2.1+, R10.1 Блок 3).
|
||||
|
||||
**v3.10** — A11 ml-ai-tooling: R10.1 Блок 3 +1 строка **Jupyter MCP** (DEFERRED — требует Python ML-окружения; ml-ai-tooling, off-phase, раздел A11 карты) + Блок 1 note (v3.10) — **promptfoo** (npm devDependency `promptfoo`, CLI-eval LLM-промптов) + **Data Scientist skill** (вендоренный сторонний скил `.claude/skills/data-scientist/`). Десятая off-phase подкатегория ml-ai-tooling. Не UI → вне R6/R14. Содержательных изменений R0–R14: 0. Связано: Tooling v2.10, Pravila v1.24, CLAUDE.md v2.10; план `docs/superpowers/plans/2026-05-17-a11-ml-ai-tooling-integration.md`.
|
||||
|
||||
**v3.9** — A3 integration-tooling: R10.1 Блок 3 +1 строка **openapi-mcp-server** (категория integration-tooling, off-phase, раздел A3 карты, stdio MCP, server `openapi` в `.mcp.json`, Tooling §4.22 #47). Не UI → вне R6/R14. Содержательных изменений R0–R14: 0. Связано: Tooling v2.9, Pravila v1.23, CLAUDE.md v2.9; план `docs/superpowers/plans/2026-05-17-a3-integration-tooling-integration.md`.
|
||||
|
||||
**v3.7** — A6-расширение deptrac: R10.1 Блок 1 +note «Блок 1 — note (v3.7)» — **deptrac** (`deptrac/deptrac` v4.6.1, Composer dev-dependency, **не** marketplace-плагин и **не** в `enabledPlugins` — регистрируется нотой, как mermaid-skill/CCPM). Категория **architecture-tooling** (Tooling #43, раздел A6 карты) — 4-й инструмент подкатегории; не UI → вне R6.0/R6.1/R14. deptrac врезан как lefthook pre-commit job 10. Содержательных изменений R0–R14: 0. Связано: Tooling v2.7, Pravila v1.21, CLAUDE.md v2.7; план `docs/superpowers/plans/2026-05-17-deptrac-architecture-fitness-integration.md`.
|
||||
@@ -418,6 +420,8 @@ Stack — **головной**. Все плагины вне stack'а — **ин
|
||||
|
||||
**Блок 1 — note (v3.7):** **deptrac** (Tooling #43, архитектурный fitness-гейт) — Composer dev-dependency (`deptrac/deptrac` v4.6.1, BSD-3), **не** marketplace-плагин и **не** в `enabledPlugins`; врезан как lefthook pre-commit job 10 (`deptrac analyse` на staged `app/**/*.php`). CLI-инструмент статического анализа направления зависимостей между слоями — не решатель; формально вне типологии трёх блоков, регистрируется здесь для полноты. Категория **architecture-tooling** (как adr-kit/architecture-patterns), вне R6.0/R6.1/R14.
|
||||
|
||||
**Блок 1 — note (v3.10):** **promptfoo** (Tooling #48, ml-ai-tooling) — npm devDependency (`promptfoo`, MIT) в корневом `package.json`, **не** marketplace-плагин и **не** в `enabledPlugins`; CLI-инструмент eval LLM-промптов, запуск `npx promptfoo` вручную/CI (платные LLM-вызовы — никогда в хук, ML1). **Data Scientist skill** (Tooling #49, ml-ai-tooling) — аналогично mermaid-skill/CCPM: вендоренный сторонний скил в `.claude/skills/data-scientist/` (`sickn33/antigravity-awesome-skills`, код MIT / контент CC BY 4.0), **не** через marketplace. Оба формально вне типологии трёх блоков, регистрируются здесь для полноты. Категория **ml-ai-tooling** (раздел A11 карты), вне R6.0/R6.1/R14.
|
||||
|
||||
**Отмена:** через удаление из `enabledPlugins` в `~/.claude/settings.json` или через live-override `/имя-плагина` (R0.4.B) на одно действие.
|
||||
|
||||
#### Блок 2: Built-in skills Claude Code (всегда доступны через `Skill` tool по `/имя`)
|
||||
@@ -450,6 +454,7 @@ Stack — **головной**. Все плагины вне stack'а — **ин
|
||||
| **Universal Icons MCP** *(`universal-icons` сервер, tools `search_icons`/`get_icon`/`health_check`)* | `.mcp.json` (`npx -y mcp-universal-icons`, MIT) | поиск/вставка SVG-иконок — 10 коллекций включая Lucide (брендовый icon-set). Категория: **design-tooling** (Tooling #45) | при поиске иконки для Vue-компонента. НЕ запрашивать jsx/Tailwind-формат (R6.0). Материал, не решатель (R10.2). Вне R14 pipeline |
|
||||
| **Figma MCP** *(remote `https://mcp.figma.com/mcp`)* — **DEFERRED** | `.mcp.json` (HTTP-транспорт, OAuth) — не установлен, precondition: Figma-аккаунт | извлечение дизайн-токенов/variables из Figma-источника (`get_variable_defs`). **Extract-only** (ADR-006) — code-gen не используется. Категория: **design-tooling** (Tooling #44) | DEFERRED (FM2 — у проекта нет Figma-файла). При появлении Figma-аккаунта. Extract-only — FD #30 остаётся UI-решателем. Вне R6.0/R6.1/R14 |
|
||||
| **openapi-mcp-server** *(`openapi` сервер, tools `mcp__openapi__*`)* | `.mcp.json` (stdio MCP, env `OPENAPI_SPEC_URL` или локальный файл) | **integration-tooling MCP** — OpenAPI/Swagger-спецификации интеграций (inspect, introspect внешних API). Категория: **integration-tooling** (Tooling §4.22 #47). Раздел A3 карты «Программирование — интеграции (API, вебхуки)». Off-phase | при работе с внешними API-интеграциями (introspection спецификаций). **READ-ONLY introspection** — не мутировать внешние API из Claude. Не trigger'ит R6.0/R6.1 фильтры и не входит в R14 pipeline UI-генераторов. Вне R6/R14 |
|
||||
| **Jupyter MCP** *(`jupyter` сервер)* — **DEFERRED** | `.mcp.json` (stdio MCP) — не установлен, precondition: Python ML-окружение | **ml-ai-tooling MCP** — исполняемые ноутбуки (классический ML: обучение моделей). Категория: **ml-ai-tooling** (Tooling §4.25 #50). Раздел A11 карты «ML / AI-разработка». Off-phase | DEFERRED — на native-Windows машине нет Python ML-рантайма и нет модели для обучения. Зарегистрирован как pending-слот (как Figma MCP); устанавливается отдельной severable-задачей при появлении конкретной модели. Вне R6/R14 |
|
||||
|
||||
**Отмена:** через удаление из `~/.claude.json` или `.mcp.json`. Live-override через `/команду` для MCP не предусмотрен — MCP-серверы не имеют slash-интерфейса.
|
||||
|
||||
|
||||
@@ -1,10 +1,12 @@
|
||||
# Правила работы Claude в проекте «Лидерра»
|
||||
|
||||
**Версия:** v1.23 (17.05.2026)
|
||||
**Версия:** v1.24 (17.05.2026)
|
||||
**Дата:** 17.05.2026
|
||||
**Назначение:** настройки проекта (Project instructions) — Claude читает этот файл в каждом чате и следует правилам ниже.
|
||||
**Статус документа:** ✅ утверждён. Содержимое скопировано в поле "Project instructions" Claude.ai. Файл хранится в архиве как служебный документ.
|
||||
|
||||
**Что изменилось в v1.24 относительно v1.23:** §13.2 +абзац «Off-phase ml-ai-tooling» — формализованы инструменты раздела A11 карты «ML / AI-разработка» (#48 promptfoo, #49 Data Scientist skill, #50 Jupyter MCP DEFERRED) как десятая off-phase подкатегория; promptfoo делает платные LLM-вызовы — только вручную/CI, никогда в хук (ML1). Границы — ADR-007. Связано: Tooling v2.10 / PSR_v1 v3.10 / CLAUDE.md v2.10. План `docs/superpowers/plans/2026-05-17-a11-ml-ai-tooling-integration.md`.
|
||||
|
||||
**Что изменилось в v1.23 относительно v1.22:** §13.2 +абзац «Off-phase integration-tooling» — формализованы инструменты раздела A3 карты «Программирование — интеграции (API, вебхуки)» (#47 openapi-mcp-server, api-docs agent) как девятая off-phase подкатегория; READ-ONLY introspection. Связано: Tooling v2.9 / PSR_v1 v3.9 / CLAUDE.md v2.9. План `docs/superpowers/plans/2026-05-17-a3-integration-tooling-integration.md`.
|
||||
|
||||
**Что изменилось в v1.22 относительно v1.21:** §13.2 +абзац «Off-phase design-tooling» — формализованы 3 инструмента раздела A4 карты «Дизайн (UI/UX, графика, бренд)» (#44 Figma MCP DEFERRED, #45 Universal Icons MCP, #46 Design plugin) как восьмая off-phase подкатегория; §13.2 PSR_v1 cross-ref синхронизирован → v3.8+. Связано: Tooling v2.8 / PSR_v1 v3.8 / CLAUDE.md v2.8. План `docs/superpowers/plans/2026-05-17-a4-design-tooling-integration.md`.
|
||||
@@ -576,6 +578,7 @@ P0 = блокер старта спринта или регуляторного
|
||||
| **v1.21** | **17.05.2026** | A6-расширение deptrac: §13.2 абзац «Off-phase architecture-tooling» расширен — формализован 4-й инструмент раздела A6 (#43 deptrac, Composer dev-dependency `deptrac/deptrac` v4.6.1 BSD-3; архитектурный fitness-гейт направления зависимостей / границ слоёв, врезан в lefthook pre-commit job 10, конфиг `app/deptrac.yaml` 13 слоёв, первый прогон 0 нарушений → baseline не нужен, red-green доказан). Категория architecture-tooling без изменений. Связано: Tooling v2.6→v2.7 (§4.18 + §0 счётчик 42→43), PSR_v1 v3.6→v3.7 (R10.1 Блок 1 note), CLAUDE.md v2.6→v2.7 (§3.3 +#43). Через manual Edit (Pravila/PSR_v1/Tooling) + `/claude-md-management:claude-md-improver` (CLAUDE.md per §5 п.10). План `docs/superpowers/plans/2026-05-17-deptrac-architecture-fitness-integration.md`. Архитектурных изменений в §§1–12 + §§13.1, 13.3–14: 0. |
|
||||
| **v1.22** | **17.05.2026** | A4 design-tooling: §13.2 +абзац «Off-phase design-tooling» — формализованы 3 инструмента раздела A4 карты «Дизайн (UI/UX, графика, бренд)» (#44 Figma MCP / #45 Universal Icons MCP / #46 Design plugin) как восьмая off-phase подкатегория, отдельная от UI-пула / infrastructure / debug-runtime / orchestration / architecture-tooling / audit-security / project-management; не UI → вне R6.0/R6.1/R14. §13.2 PSR_v1 cross-ref v3.3+ → v3.8+ (текст застрял на v3.3+ — changelog v1.18-v1.20 заявлял bump'ы, но §13.2 не обновлялся; теперь синхронизирован). Связано: Tooling v2.8 / PSR_v1 v3.8 / CLAUDE.md v2.8. План `docs/superpowers/plans/2026-05-17-a4-design-tooling-integration.md`. |
|
||||
| **v1.23** | **17.05.2026** | A3 integration-tooling: §13.2 +абзац «Off-phase integration-tooling» — формализованы инструменты раздела A3 карты «Программирование — интеграции (API, вебхуки)» (#47 `openapi-mcp-server`, Tooling §4.22; `api-docs` agent, claude-flow, без Tooling-номера) как девятая off-phase подкатегория, отдельная от всех предыдущих; не UI → вне R6.0/R6.1/R14. READ-ONLY introspection. Регулируются PSR_v1 R10.1 Блок 3. Связано: Tooling v2.9 / PSR_v1 v3.9 / CLAUDE.md v2.9. План `docs/superpowers/plans/2026-05-17-a3-integration-tooling-integration.md`. Архитектурных изменений в §§1–12 + §§13.1, 13.3–14: 0. |
|
||||
| **v1.24** | **17.05.2026** | A11 ml-ai-tooling: §13.2 +абзац «Off-phase ml-ai-tooling» — формализованы инструменты раздела A11 карты «ML / AI-разработка» (#48 promptfoo — npm devDependency, CLI-eval LLM-промптов; #49 Data Scientist skill — вендоренный сторонний скил; #50 Jupyter MCP — DEFERRED, требует Python ML-окружения) как десятая off-phase подкатегория, отдельная от всех предыдущих; не UI → вне R6.0/R6.1/R14. promptfoo делает платные LLM-вызовы — только вручную/CI, никогда в хук (ML1). Границы — ADR-007. Связано: Tooling v2.10 / PSR_v1 v3.10 / CLAUDE.md v2.10. Через manual Edit всех 4 нормативных файлов (claude-md-management неприменим — исполнение в worktree, §5 п.10 worktree-constraint эксцепшн). План `docs/superpowers/plans/2026-05-17-a11-ml-ai-tooling-integration.md`. Архитектурных изменений в §§1–12 + §§13.1, 13.3–14: 0. |
|
||||
|
||||
---
|
||||
|
||||
@@ -720,6 +723,8 @@ Frontend Design и `obra/superpowers` (v5.1.0, 14 skills) — **парный sta
|
||||
|
||||
**Off-phase integration-tooling (A3):** Инструменты раздела A3 карты «Программирование — интеграции (API, вебхуки)» — #47 `openapi-mcp-server` (Tooling §4.22; введён A3-интеграцией 17.05.2026) и `api-docs` agent (claude-flow, узел карты A3 без отдельного Tooling-номера). Off-phase, не UI → вне R6/R14 PSR_v1. READ-ONLY introspection. Регулируются PSR_v1 R10.1 Блок 3.
|
||||
|
||||
**Off-phase ml-ai-tooling (A11, v1.24, 17.05.2026):** Инструменты раздела A11 карты «ML / AI-разработка» — #48 `promptfoo` (Tooling §4.23; npm devDependency, CLI-eval LLM-промптов, MIT), #49 `Data Scientist skill` (Tooling §4.24; вендоренный сторонний скил в `.claude/skills/data-scientist/`, классический ML-воркфлоу; код MIT / контент CC BY 4.0), #50 `Jupyter MCP` (Tooling §4.25; **DEFERRED** — требует Python ML-окружения, на native-Windows машине не ставится; зарегистрирован как pending-слот, как Figma MCP #44). Плюс reuse-слой — claude-api skill (PSR_v1 R10.1 Блок 2), context7 MCP, Sentry MCP — без новых номеров. Десятая off-phase подкатегория. Off-phase, не UI → вне R6.0/R6.1/R14 PSR_v1. promptfoo делает платные LLM-вызовы — запуск только вручную/CI, никогда в хук (конфликт-аудит ML1). Границы — ADR-007. Регулируются PSR_v1 R10.1 (Блок 1 — promptfoo dev-dep + Data Scientist skill вендорен; Блок 3 — Jupyter MCP). Установлены 17.05.2026 на ветке `worktree-a11-ml-ai-tooling`; план `docs/superpowers/plans/2026-05-17-a11-ml-ai-tooling-integration.md`.
|
||||
|
||||
### 13.3. Скоуп
|
||||
|
||||
| Тип задачи | Кто отвечает |
|
||||
|
||||
+49
-5
File diff suppressed because one or more lines are too long
@@ -2,6 +2,7 @@
|
||||
|
||||
- **Status:** Accepted
|
||||
- **Date:** 2026-05-17
|
||||
- **Amended:** 2026-05-17 — Decision item 4 added (Universal Icons icon-path boundary).
|
||||
- **Deciders:** Дмитрий
|
||||
|
||||
## Context
|
||||
@@ -29,13 +30,27 @@ Figma account yet); the boundary still applies the moment it is connected.
|
||||
Phase-8 review stays with the PSR_v1 R5 aspect-split (FD owns the UI/UX aspect)
|
||||
plus the Superpowers review skills. The Design plugin does not replace
|
||||
`superpowers:requesting-code-review`.
|
||||
4. **Universal Icons MCP raw-SVG is for non-Lucide collections only** (amendment
|
||||
2026-05-17). Lucide is the project's branded icon set (CTO-19), rendered via the
|
||||
`lucide-vue-next` component package plus the custom Vuetify `IconSet` mapping in
|
||||
`app/resources/js/plugins/vuetify.ts` (103-entry map). For any Lucide icon that
|
||||
component path is canonical. Universal Icons MCP `get_icon` raw-SVG output is
|
||||
used only for collections `lucide-vue-next` does not provide (Heroicons, Tabler,
|
||||
Phosphor, etc.), and the SVG is wrapped into a Vue component — never inlined to
|
||||
bypass the icon system. ADR-006 originally regulated #45 only against 21st
|
||||
`logo_search`; this item closes the previously unregulated #45 ↔
|
||||
`lucide-vue-next` boundary.
|
||||
|
||||
## Consequences
|
||||
|
||||
- A Figma MCP code-generation call is a process violation (CLAUDE.md §5 п.6).
|
||||
- Universal Icons (#45) covers UI icons; 21st `logo_search` covers brand logos —
|
||||
distinct, both retained.
|
||||
- These boundaries are mirrored as PSR_v1 R10.1 rows + R6/R10/R14 notes.
|
||||
- Pulling a Lucide icon as raw SVG via Universal Icons MCP, instead of
|
||||
`lucide-vue-next`, is a process violation (CLAUDE.md §5 п.6 — two tools on one
|
||||
task).
|
||||
- These boundaries are mirrored as PSR_v1 R10.1 rows + R6/R10/R14 notes; the
|
||||
Decision-4 icon-path boundary is mirrored in CLAUDE.md §3.3 #45 and Tooling §4.20.
|
||||
|
||||
## Enforcement
|
||||
|
||||
|
||||
@@ -0,0 +1,39 @@
|
||||
# ADR-007: ML / AI tooling (A11)
|
||||
|
||||
- **Status:** Accepted
|
||||
- **Date:** 2026-05-17
|
||||
- **Deciders:** Дмитрий
|
||||
|
||||
## Context
|
||||
|
||||
The `A11 «ML / AI-разработка»` map section had zero tooling. Лидерра ships no
|
||||
ML/AI code; `calc_lead_score` is a deterministic SQL function. A toolset is needed
|
||||
for the day AI features (LLM-backed) or a scoring model are scoped.
|
||||
|
||||
## Decision
|
||||
|
||||
A11 adopts a six-position toolset in two subcategories:
|
||||
|
||||
- **LLM integration** — the claude-api skill (build), promptfoo (test prompts),
|
||||
Sentry MCP (observe). All reuse or new-light.
|
||||
- **Classical ML** — a vendored Data Scientist skill (workflow knowledge). The
|
||||
executable part, **Jupyter MCP**, is **deferred**: it needs a Python ML runtime
|
||||
the deliberately-minimal native-Windows machine lacks, and there is no model to
|
||||
train. Jupyter MCP is a reserved registry slot, installed by a separate task
|
||||
when a concrete model is scoped.
|
||||
- promptfoo runs manually / CI only — never in a hook (paid LLM calls).
|
||||
- A11 tools are non-UI → the `ml-ai-tooling` off-phase category.
|
||||
|
||||
## Consequences
|
||||
|
||||
- Positive: A11 populated; AI features have a build+test+observe toolchain.
|
||||
- Risk: the Data Scientist skill is third-party (CC BY 4.0 content) — mitigated by
|
||||
vendoring with attribution into `.claude/skills/data-scientist/`.
|
||||
- Cost: promptfoo is a heavy devDependency (~1090 transitive packages, one native
|
||||
module). Accepted — it is dev-only tooling, not shipped to the app.
|
||||
- Deferred: no Python runtime until a model is scoped — accepted, this is the
|
||||
decision.
|
||||
|
||||
## Enforcement
|
||||
|
||||
None — A11 tools are advisory; verified by use and code review.
|
||||
+53
-11
@@ -228,10 +228,10 @@ function pos(ring, angleDeg) {
|
||||
|
||||
const NODES = [
|
||||
// ── ПРАВИЛА (4) ── центр + первое кольцо ───────
|
||||
{ id: 'pravila', label: 'Pravila v1.16', group: 'rules', size: 38, ring: 0, ...pos(0, 0) },
|
||||
{ id: 'claude_md', label: 'CLAUDE.md v2.2', group: 'rules', size: 34, ring: 1, ...pos(1, 30) },
|
||||
{ id: 'psr_v1', label: 'PSR_v1 v3.2', group: 'rules', size: 32, ring: 1, ...pos(1, 150) },
|
||||
{ id: 'tooling', label: 'Tooling v2.2', group: 'rules', size: 30, ring: 1, ...pos(1, 270) },
|
||||
{ id: 'pravila', label: 'Pravila v1.24', group: 'rules', size: 38, ring: 0, ...pos(0, 0) },
|
||||
{ id: 'claude_md', label: 'CLAUDE.md v2.10', group: 'rules', size: 34, ring: 1, ...pos(1, 30) },
|
||||
{ id: 'psr_v1', label: 'PSR_v1 v3.10', group: 'rules', size: 32, ring: 1, ...pos(1, 150) },
|
||||
{ id: 'tooling', label: 'Tooling v2.10', group: 'rules', size: 30, ring: 1, ...pos(1, 270) },
|
||||
|
||||
// ── ПЛАГИНЫ (13) ── второе кольцо ──────────────
|
||||
{ id: 'superpowers', label: 'Superpowers v5.1', group: 'plugins', size: 30, ring: 2, ...pos(2, 45) },
|
||||
@@ -282,6 +282,10 @@ const NODES = [
|
||||
{ id: 'sk_audit_portal', label: 'audit-portal', group: 'skills_proj', size: 20, ring: 3, ...pos(3, 325) },
|
||||
// C9 project-management-tooling (17.05.2026) — вендоренный скил раздела «Управление проектами»
|
||||
{ id: 'ccpm', label: 'CCPM\n(skill)', group: 'skills_proj', size: 18, ring: 3, ...pos(3, 335) },
|
||||
// A11 ml-ai-tooling (17.05.2026) — скилы и CLI раздела «ML / AI-разработка»
|
||||
{ id: 'claude_api', label: 'claude-api\n(skill)', group: 'skills_proj', size: 18, ring: 3, ...pos(3, 345) },
|
||||
{ id: 'data_scientist', label: 'Data Scientist\n(skill)', group: 'skills_proj', size: 18, ring: 3, ...pos(3, 355) },
|
||||
{ id: 'promptfoo', label: 'promptfoo', group: 'plugins', size: 20, ring: 2, ...pos(2, 365) },
|
||||
|
||||
// ── ХУКИ (12) — S+infra + E (economy/skill) ───
|
||||
{ id: 'hk_session', label: 'SessionStart:\ncontext-inject', group: 'hooks', size: 24, ring: 4, ...pos(4, 100) },
|
||||
@@ -542,6 +546,11 @@ const EDGES = [
|
||||
E('tooling', 'mcp_openapi', '§4.22 #47 — реестр'),
|
||||
E('ag_apidocs', 'mcp_openapi', 'спека → MCP-ресурс'),
|
||||
|
||||
// ── A11 ML-AI-TOOLING 17.05.2026 — связи новых узлов ──
|
||||
E('psr_v1', 'promptfoo', 'R10.1 блок 1:\nml-ai-tooling'),
|
||||
E('tooling', 'claude_api', 'reuse — built-in skill\n(PSR_v1 R10.1 блок 2)'),
|
||||
E('tooling', 'data_scientist', '§4.24 #49 — реестр'),
|
||||
|
||||
// ══════════════════════════════════════════════════
|
||||
// КОНФЛИКТЫ — 3-color classification (iter2 §4)
|
||||
// 🔴 не закрыт правилом / ⚫ возник на практике / 🟢 закрыт правилом
|
||||
@@ -632,7 +641,7 @@ const NODE_DETAILS = {
|
||||
'Править можно только через скил `/claude-md-management:claude-md-improver` или `:revise-claude-md` (правило §5 п.10). Прямые Edit/Write блокируются хуком предупреждения.',
|
||||
[{ name: 'Pravila', cond: 'всегда подчинён (уровень 2a)' }],
|
||||
[
|
||||
{ name: 'Tooling v2.2', cond: 'ссылается как на реестр инструментов' },
|
||||
{ name: 'Tooling v2.10', cond: 'ссылается как на реестр инструментов' },
|
||||
{ name: 'плагин claude-md-management', cond: 'правило §5 п.10 — единственный канал правок' }
|
||||
],
|
||||
[
|
||||
@@ -655,7 +664,7 @@ const NODE_DETAILS = {
|
||||
[{ name: 'CLAUDE.md', desc: 'CLAUDE.md §5 п.10 требует править только через скил claude-md-management, а PSR_v1 это ограничение не повторяет — риск прямых Edit', type: 'GREEN' }]
|
||||
),
|
||||
tooling: nd(
|
||||
'Реестр 55 позиций — 35 формализованных инструментов + 20 ruflo-плагинов; §4.10 — ruflo как advisory/automation-подсистема. Когда что использовать, команды установки, конфликты.',
|
||||
'Реестр 70 позиций — 50 формализованных инструментов + 20 ruflo-плагинов; §4.10 — ruflo как advisory/automation-подсистема. Когда что использовать, команды установки, конфликты.',
|
||||
'При выборе инструмента для фазы (нулевая документация / первая backend / вторая frontend / третья перед запуском в боевую среду), при добавлении нового инструмента, при обновлении версий.',
|
||||
'При прямом конфликте с CLAUDE.md побеждает CLAUDE.md (оперативная карта уровня 2a). Любая правка требует синхронизации с CLAUDE.md §3.',
|
||||
[
|
||||
@@ -851,6 +860,32 @@ const NODE_DETAILS = {
|
||||
[{ name: 'docs/api/', cond: 'источник OpenAPI-спеки' }]
|
||||
),
|
||||
|
||||
// ── A11 ML-AI-TOOLING (17.05.2026) ──────────────
|
||||
claude_api: nd(
|
||||
'Скил сборки AI-фич на Anthropic SDK (prompt-кэш). Reuse — раздел A11 опирается также на context7 MCP (доки) и Sentry MCP (LLM-наблюдаемость).',
|
||||
'При разработке AI-фич на Anthropic API / Claude SDK — скил задаёт паттерны prompt-кэша, batch-запросов, tool use.',
|
||||
'Reuse-узел раздела A11 (ml-ai-tooling). claude-api — встроенный скил Claude Code, не нумерованная Tooling-позиция; регистрация — Tooling «built-in skills» + PSR_v1 R10.1 блок 2. В A11 — reuse-слой (CLAUDE.md §6). Не UI → вне фильтров R6.0/R6.1/R14.',
|
||||
[{ name: 'Tooling', cond: 'built-in skill — PSR_v1 R10.1 блок 2 (reuse)' }],
|
||||
[],
|
||||
[{ name: 'context7 MCP', cond: 'документация Anthropic SDK' }, { name: 'Sentry MCP', cond: 'LLM-наблюдаемость (off-phase reuse)' }]
|
||||
),
|
||||
promptfoo: nd(
|
||||
'npm-CLI eval LLM-промптов: ассерты, регрессия, red-team. Запуск вручную/CI — не в хуках (платные LLM-вызовы).',
|
||||
'При разработке и проверке AI-промптов — запуск test-suite promptfoo вручную или в CI. Никогда в pre-commit хук (ML1: платные вызовы).',
|
||||
'Правило PSR_v1 R10.1 блок 1 note (ml-ai-tooling, off-phase). npm devDependency, тяжёлый (~1090 пакетов). Не UI → вне фильтров R6.0/R6.1/R14. Tooling §4.23 #48, CLAUDE.md §3.3 #48.',
|
||||
[{ name: 'PSR_v1', cond: 'R10.1 блок 1: ml-ai-tooling' }],
|
||||
[{ name: 'ML1', cond: 'никогда в хук/pre-commit — платные LLM-вызовы' }],
|
||||
[]
|
||||
),
|
||||
data_scientist: nd(
|
||||
'Vendored-скил: классический ML-воркфлоу — выбор алгоритма, feature engineering, оценка модели.',
|
||||
'При ML-задаче (выбор алгоритма, feature engineering, валидация модели) — knowledge-only playbook без генерации кода.',
|
||||
'Вендорен в .claude/skills/data-scientist/ (ML3 — lefthook markdownlint+cspell исключают через job-exclude). Knowledge-only, не решатель. Не UI → вне фильтров R6.0/R6.1/R14. Tooling §4.24 #49, CLAUDE.md §3.3 #49.',
|
||||
[{ name: 'Tooling', cond: '§4.24 #49 — реестр' }],
|
||||
[],
|
||||
[]
|
||||
),
|
||||
|
||||
// ── СКИЛЫ SUPERPOWERS ────────────────────────────
|
||||
sk_brainstorm: nd(
|
||||
'Продумывает задачу вместе с заказчиком, формулирует варианты A/B/C и согласует дизайн до написания кода.',
|
||||
@@ -1794,10 +1829,10 @@ const META_WINDOW = '09–16.05.2026'; // окно подсчёта испо
|
||||
// usesSrc: 'скил' | 'агент' | 'MCP' | 'хук' | 'memory-чтение' | 'коммиты' | 'инспекция' | '—'
|
||||
const NODE_META = {
|
||||
// ── ПРАВИЛА (4) — узлы-правила, напрямую не вызываются ──
|
||||
pravila: { since: '06.05.2026', changed: '16.05.2026', uses: null, usesSrc: '—' },
|
||||
claude_md: { since: '06.05.2026', changed: '16.05.2026', uses: null, usesSrc: '—' },
|
||||
psr_v1: { since: '09.05.2026', changed: '16.05.2026', uses: null, usesSrc: '—' },
|
||||
tooling: { since: '06.05.2026', changed: '16.05.2026', uses: null, usesSrc: '—' },
|
||||
pravila: { since: '06.05.2026', changed: '17.05.2026', uses: null, usesSrc: '—' },
|
||||
claude_md: { since: '06.05.2026', changed: '17.05.2026', uses: null, usesSrc: '—' },
|
||||
psr_v1: { since: '09.05.2026', changed: '17.05.2026', uses: null, usesSrc: '—' },
|
||||
tooling: { since: '06.05.2026', changed: '17.05.2026', uses: null, usesSrc: '—' },
|
||||
|
||||
// ── ПЛАГИНЫ (5) ──
|
||||
superpowers: { since: '09.05.2026', changed: '—', uses: null, usesSrc: '—' },
|
||||
@@ -1948,6 +1983,11 @@ const NODE_META = {
|
||||
// ── A3 INTEGRATION-TOOLING (17.05.2026) ──
|
||||
ag_apidocs: { since: '17.05.2026', changed: '—', uses: null, usesSrc: '—' },
|
||||
mcp_openapi: { since: '17.05.2026', changed: '—', uses: null, usesSrc: '—' },
|
||||
|
||||
// ── A11 ML-AI-TOOLING (17.05.2026) ──
|
||||
claude_api: { since: '17.05.2026', changed: '—', uses: null, usesSrc: 'скил' },
|
||||
promptfoo: { since: '17.05.2026', changed: '—', uses: null, usesSrc: 'CLI' },
|
||||
data_scientist: { since: '17.05.2026', changed: '—', uses: null, usesSrc: 'скил' },
|
||||
};
|
||||
|
||||
// Явные парные дубли (Фича 3) — попадают в кнопку «⧉ Дубли».
|
||||
@@ -2030,7 +2070,7 @@ const SECTIONS = [
|
||||
{ id: 'E7', bucket: 'E', label: 'Исследования' },
|
||||
{ id: 'E8', bucket: 'E', label: 'Самообучение Claude' },
|
||||
];
|
||||
// Узел -> раздел. Покрывает все 118 узлов карты.
|
||||
// Узел -> раздел. Покрывает все 121 узлов карты.
|
||||
const NODE_SECTION = {
|
||||
// правила (4)
|
||||
pravila: 'E1', claude_md: 'E1', psr_v1: 'E1', tooling: 'E1',
|
||||
@@ -2082,6 +2122,8 @@ const NODE_SECTION = {
|
||||
mcp_figma: 'A4', mcp_icons: 'A4', design_plugin: 'A4',
|
||||
// A3 integration-tooling 17.05.2026 — раздел «Программирование — интеграции» наполнен
|
||||
ag_apidocs: 'A3', mcp_openapi: 'A3',
|
||||
// A11 ml-ai-tooling 17.05.2026 — раздел «ML / AI-разработка» наполнен
|
||||
claude_api: 'A11', promptfoo: 'A11', data_scientist: 'A11',
|
||||
};
|
||||
// Вторичная классификация: узел первично в NODE_SECTION, дополнительно — в этих
|
||||
// разделах (кросс-реф). Введено A3-интеграцией 17.05.2026 — раздел A3 наполняется
|
||||
|
||||
@@ -0,0 +1,48 @@
|
||||
# docs/ml — ML / AI playbook (map section A11)
|
||||
|
||||
Home of the `A11 «ML / AI-разработка»` section. Defines the tooling Лидерра uses
|
||||
to build and test ML/AI capability. The portal currently ships no ML/AI code —
|
||||
this section is the toolset, ready for when AI features are scoped.
|
||||
|
||||
## Toolset
|
||||
|
||||
| Tool | Role | Status |
|
||||
|---|---|---|
|
||||
| **claude-api skill** | Build AI features on the Anthropic SDK (lead qualification, call summaries, email drafts) with prompt caching. | reuse — already available |
|
||||
| **context7 MCP** | Up-to-date docs for AI/ML libraries and SDKs. | reuse — already installed |
|
||||
| **Sentry MCP** | Debug AI features in production via Sentry AI/LLM monitoring (read-only). | reuse — Tooling #34, pending the Sentry deployment (Б-1) |
|
||||
| **promptfoo** | Test suite for LLM prompts/agents: assertions, regression, LLM-graded eval, red-team. | installed — `npx promptfoo` |
|
||||
| **Data Scientist skill** | Classical-ML workflow: business objective → ML task, algorithm selection, feature engineering, evaluation. | installed — vendored skill |
|
||||
| **Jupyter MCP** | Executable notebooks for real model training. | **deferred** — see below |
|
||||
|
||||
## Boundaries (which tool for which job)
|
||||
|
||||
- **Building an AI feature** (a prompt-backed endpoint) → the **claude-api skill**.
|
||||
- **Testing / regression-checking an LLM prompt** → **promptfoo** (`docs/ml/promptfoo-example/`).
|
||||
- **A classical-ML modelling question** (which algorithm, how to evaluate) → the
|
||||
**Data Scientist skill** (`.claude/skills/data-scientist/`).
|
||||
- **Executing a notebook / training a model** → **Jupyter MCP** — *deferred*.
|
||||
- promptfoo's **red-team** tests *prompts*; the D3 Trail of Bits / Semgrep tools do
|
||||
SAST of *code*. Different objects — not a duplication.
|
||||
|
||||
## promptfoo — running an eval
|
||||
|
||||
promptfoo makes **paid** Anthropic API calls. It runs **manually or in CI only** —
|
||||
never in a git hook, never in pre-commit, never automatically.
|
||||
|
||||
- API key: `ANTHROPIC_API_KEY` env var (PowerShell User scope — the Sentry
|
||||
`SENTRY_AUTH_TOKEN` pattern). Never commit a key.
|
||||
- Run the seed example: `npm run eval:llm` (or
|
||||
`npx promptfoo eval -c docs/ml/promptfoo-example/promptfooconfig.yaml`).
|
||||
- Footprint note: promptfoo is a large devDependency (~1090 transitive packages,
|
||||
one native module — `better-sqlite3` — which `prebuild-install` fetches as a
|
||||
prebuilt binary; no local C++ toolchain is required when the prebuild download
|
||||
succeeds). It is dev-tooling only — not shipped to the Laravel app.
|
||||
|
||||
## Jupyter MCP — why deferred
|
||||
|
||||
Jupyter MCP executes notebooks; it needs a Python ML environment (pandas /
|
||||
scikit-learn / Jupyter). The machine is native Windows, deliberately runtime-minimal
|
||||
(no Docker), and there is no model to train yet. Jupyter MCP is a **reserved slot**:
|
||||
registered in the Tooling registry as *pending*, installed by a separate severable
|
||||
task when a concrete ML model is scoped. See the A11 plan's "Deferred Task".
|
||||
@@ -0,0 +1,20 @@
|
||||
# promptfoo example — lead-qualification eval
|
||||
|
||||
A worked promptfoo eval: a HOT/WARM/COLD lead-classification prompt with three
|
||||
assertion cases. Demonstrates the A11 prompt-testing workflow.
|
||||
|
||||
## Run
|
||||
|
||||
```bash
|
||||
# from the repo root; ANTHROPIC_API_KEY must be set (PowerShell User scope)
|
||||
npm run eval:llm
|
||||
```
|
||||
|
||||
This makes **paid** Anthropic API calls. Run it manually or in CI only — never
|
||||
in a git hook or pre-commit (A11 rule ML1). See `docs/ml/README.md`.
|
||||
|
||||
## Adapt
|
||||
|
||||
Copy `promptfooconfig.yaml` next to a real prompt when an AI feature is built.
|
||||
Swap the model, add `tests`, use richer assertions (`contains`, `llm-rubric`,
|
||||
cost/latency thresholds). Full reference: <https://promptfoo.dev/docs/>.
|
||||
@@ -0,0 +1,31 @@
|
||||
# yaml-language-server: $schema=https://promptfoo.dev/config-schema.json
|
||||
# Seed example — A11. Lead-qualification prompt eval.
|
||||
# Run manually: npm run eval:llm (needs ANTHROPIC_API_KEY — never in CI/hooks)
|
||||
description: "Лидерра — lead-qualification prompt eval (example)"
|
||||
|
||||
prompts:
|
||||
- |
|
||||
Классифицируй обращение лида как HOT, WARM или COLD.
|
||||
Ответь РОВНО одним словом — HOT, WARM или COLD.
|
||||
|
||||
Обращение: {{message}}
|
||||
|
||||
providers:
|
||||
- id: anthropic:messages:claude-haiku-4-5-20251001
|
||||
|
||||
tests:
|
||||
- vars:
|
||||
message: "Нужно срочно, бюджет согласован, готовы подписать договор сегодня."
|
||||
assert:
|
||||
- type: equals
|
||||
value: HOT
|
||||
- vars:
|
||||
message: "Интересно, расскажите подробнее про условия и сроки."
|
||||
assert:
|
||||
- type: equals
|
||||
value: WARM
|
||||
- vars:
|
||||
message: "Просто смотрю что есть на рынке, ничего конкретного."
|
||||
assert:
|
||||
- type: equals
|
||||
value: COLD
|
||||
+4
-2
@@ -31,11 +31,12 @@ pre-commit:
|
||||
# 2. markdownlint — стиль Markdown с авто-fix
|
||||
- name: markdownlint
|
||||
glob: "*.md"
|
||||
# Вендоренные сторонние скилы .claude/skills/{mermaid,ccpm}/ — не линтуем (их .md
|
||||
# Вендоренные сторонние скилы .claude/skills/{mermaid,ccpm,data-scientist}/ — не линтуем (их .md
|
||||
# не наши; markdownlint-cli2 игнорирует .markdownlintignore при явных путях).
|
||||
exclude:
|
||||
- ".claude/skills/mermaid/**"
|
||||
- ".claude/skills/ccpm/**"
|
||||
- ".claude/skills/data-scientist/**"
|
||||
run: npx markdownlint-cli2 --fix {staged_files}
|
||||
stage_fixed: true
|
||||
fail_text: |
|
||||
@@ -47,10 +48,11 @@ pre-commit:
|
||||
# (useGitignore:true) игнорирует worktree-коммиты под gitignored .claude/worktrees/.
|
||||
- name: cspell
|
||||
glob: "*.md"
|
||||
# Вендоренные сторонние скилы .claude/skills/{mermaid,ccpm}/ — не проверяем орфографию.
|
||||
# Вендоренные сторонние скилы .claude/skills/{mermaid,ccpm,data-scientist}/ — не проверяем орфографию.
|
||||
exclude:
|
||||
- ".claude/skills/mermaid/**"
|
||||
- ".claude/skills/ccpm/**"
|
||||
- ".claude/skills/data-scientist/**"
|
||||
run: npx cspell --no-progress --no-summary --no-gitignore {staged_files}
|
||||
fail_text: |
|
||||
cspell нашёл слова, отсутствующие в словаре.
|
||||
|
||||
Generated
+10927
-3
File diff suppressed because it is too large
Load Diff
+3
-1
@@ -15,7 +15,8 @@
|
||||
"a11y": "pa11y-ci --config pa11y.config.json",
|
||||
"a11y:handoff": "pa11y-ci --config pa11y-handoff.config.json",
|
||||
"check:docs": "run-p lint:md spell links a11y",
|
||||
"sast": "semgrep --config=p/php --config=p/javascript --config=p/typescript --config=p/secrets --config=.semgrep.yml --error --time"
|
||||
"sast": "semgrep --config=p/php --config=p/javascript --config=p/typescript --config=p/secrets --config=.semgrep.yml --error --time",
|
||||
"eval:llm": "promptfoo eval -c docs/ml/promptfoo-example/promptfooconfig.yaml"
|
||||
},
|
||||
"devDependencies": {
|
||||
"@cspell/dict-en_us": "^4.4.33",
|
||||
@@ -27,6 +28,7 @@
|
||||
"pa11y": "^9.1.1",
|
||||
"pa11y-ci": "^4.1.0",
|
||||
"postcss-html": "^1.8.1",
|
||||
"promptfoo": "^0.121.11",
|
||||
"stylelint": "^17.11.0",
|
||||
"stylelint-config-standard": "^40.0.0"
|
||||
},
|
||||
|
||||
Reference in New Issue
Block a user