agentspec-copilot

Create a pull request with conventional commits and a structured description. Supports GitHub and Azure DevOps — platform is auto-detected from the git remote. Use when the user wants to open a PR, commit changes, or submit work for review.

Create a pull request with conventional commits and a structured description. Supports GitHub and Azure DevOps — platform is auto-detected from the git remote. Use when the user wants to open a PR, commit changes, or submit work for review.

Analyze meeting transcripts — extract decisions, action items, and create SSOT documents. Use when processing standup notes, stakeholder reviews, or sprint meetings.

Save valuable insights from the current session to storage. Use when you've discovered decisions, patterns, or gotchas worth remembering across sessions.

Generate comprehensive, production-ready README.md by analyzing codebase with explorer + documenter agents. Use when creating or updating project documentation.

Sync project context to copilot-instructions.md by analyzing codebase patterns and conventions. Use after adding agents, commands, or significant architecture changes.

RAG/embedding pipeline scaffolding — delegates to de-ai-data-engineer agent. Use when building RAG pipelines, embedding workflows, feature stores, or text- to-SQL systems.

Data contract authoring (ODCS) — delegates to test-data-contracts-engineer agent. Use when defining producer-consumer agreements, SLAs, or schema governance rules.

Data quality rules generation — delegates to test-data-quality-analyst agent. Use when generating Great Expectations suites, dbt tests, or SQL quality assertions for your data models.

Table format and catalog guidance — delegates to architect-lakehouse agent. Use when working with Iceberg, Delta Lake, catalogs, or migrating from legacy table formats.

Legacy ETL migration — delegates to de-dbt-specialist and de-spark-engineer agents. Use when converting stored procedures, SSIS packages, or Informatica workflows to modern stack.

DAG/pipeline scaffolding — delegates to architect-pipeline agent. Use when scaffolding Airflow or Dagster pipelines with best-practice patterns.

Interactive schema design — delegates to architect-schema-designer agent. Use when designing star schemas, Data Vault models, SCD strategies, or DDL for any target platform.

SQL-specific code review — delegates to de-sql-optimizer + python-code- reviewer agents. Use when reviewing SQL for anti-patterns, performance issues, or cross-dialect correctness.

Exploratory data analysis for data scientists — delegates to ds-eda-analyst agent. Use when profiling datasets, visualizing distributions, checking correlations, or detecting missing data and outliers.

Experiment tracking and ML deployment for data scientists — delegates to ds- experiment-tracker and ds-ml-deployer agents. Use when logging MLflow runs, comparing experiments, registering models, promoting to production, serving via API, or running batch inference.

Feature engineering for data scientists — delegates to ds-feature-engineer agent. Use when building scikit-learn Pipelines, encoding categoricals, imputing missing values, scaling features, or creating lag/rolling features for time series.

Model evaluation for data scientists — delegates to ds-model-evaluator agent. Use when computing classification or regression metrics, plotting ROC/PR curves, calibrating probabilities, tuning decision thresholds, or comparing models statistically.

Model training for data scientists — delegates to ds-model-trainer agent. Use when training classification or regression models, tuning hyperparameters with Optuna, running cross-validation, or comparing multiple algorithms.

Create Excalidraw diagram JSON files that make visual arguments. Use when the user wants to visualize workflows, architectures, or concepts.

Create a complete KB domain from scratch with MCP validation. Use when adding a new knowledge base domain for a technology, library, or concept.

Dual AI code review with CodeRabbit + Claude Code for maximum coverage. Use when reviewing changes before a PR, or for deep architectural analysis of committed code.

Generate a visual HTML diff review — before/after architecture comparison with code review analysis. Use when visualizing code changes, comparing branches, or reviewing PRs visually.

Verify the factual accuracy of a document against the actual codebase, correct inaccuracies in place. Use when validating HTML review pages, plan documents, or any document making claims about code.

Generate a stunning magazine-quality slide deck as a self-contained HTML page. Use when creating presentations, stakeholder decks, or visual storytelling for any technical topic.

Generate a visual HTML implementation plan — detailed feature specification with state machines, code snippets, and edge cases. Use when planning a feature implementation with architectural depth.

Generate a beautiful standalone HTML diagram and open it in the browser. Use when visualizing workflows, architectures, systems, or any concept that benefits from a rich visual page.

Generate a visual HTML plan review — current codebase state vs. proposed implementation plan. Use when evaluating a spec, RFC, or design document against the actual codebase.

Generate a visual HTML project recap — rebuild mental model of a project's current state, recent decisions, and cognitive debt hotspots. Use when re- entering a project or summarizing recent activity.

Share a visual explainer HTML file instantly via Vercel — returns a live URL with no authentication required. Use when sharing diagrams, reviews, or slide decks with others.

Generate beautiful, self-contained HTML pages that visually explain systems, code changes, plans, and data. Use when the user asks for a diagram, architecture overview, diff review, plan review, project recap, comparison table, or any visual explanation of technical concepts. Also use proactively when you are about to render a complex ASCII table (4+ rows or 3+ columns) — present it as a styled HTML page instead. license: MIT compatibility: Requires a browser to view generated HTML files. Optional surf-cli for AI image generation. metadata: author: nicobailon version: "0.5.1"

Explore ideas through collaborative dialogue before requirements capture (Phase 0). Use when the user wants to brainstorm a feature, explore an idea, or hasn't defined requirements yet.

Execute implementation with on-the-fly task generation from a DESIGN document (Phase 3). Use when the user wants to implement a feature that has a completed DESIGN document.

Capture and validate requirements in one pass (Phase 1). Use when the user wants to extract requirements from notes, emails, a BRAINSTORM document, or a direct request.

Create architecture and technical specification from a DEFINE document (Phase 2). Use when the user wants to design a solution, create an architecture, or plan implementation after requirements are captured.

Update any SDD phase document when requirements or design change (Cross- Phase). Use when the user needs to modify a BRAINSTORM, DEFINE, or DESIGN document mid-stream and wants cascade awareness.

Archive a completed feature with lessons learned (Phase 4). Use when the build is done, all tests pass, and the user wants to finalize and archive the feature.

Sync project context to copilot-instructions.md by analyzing codebase patterns and conventions. Use after adding agents, commands, or significant architecture changes.

Create Excalidraw diagram JSON files that make visual arguments. Use when the user wants to visualize workflows, architectures, or concepts.

Create a complete KB domain from scratch with MCP validation. Use when adding a new knowledge base domain for a technology, library, or concept.