Skills
Migrate Metabase to Bonnard Semantic Layer
This skill guides the development of a semantic layer by analyzing an existing Metabase instance. It walks the user through connecting to Metabase, generating an intelligence report, and subsequently building structured cubes and views usin…
Semantic Layer Setup and Deployment Guide
This skill guides users through the end-to-end process of building and deploying a semantic layer using the Bonnard CLI. It facilitates connecting data sources, defining structured cubes with measures and dimensions, and creating curated vi…
Semantic Layer Design Guide for AI Agents
This guide outlines best practices for designing semantic data layers, ensuring views are audience-centric and optimized for consumption by both human users and AI agents. It covers principles like starting from user questions, implementing…
Build and Deploy Data Dashboards with Markdown
This skill guides users through creating interactive data dashboards using a markdown format, enabling the definition of queries and visualization components. It facilitates local previewing and deployment to the Bonnard platform via dedica…
Mémoire Design Intelligence and Tooling
Provides a local design intelligence layer for managing design systems, Figma context, and research synthesis. It enables automated UI audits, component specification generation, and shadcn/ui code generation using Atomic Design principles.
Mémoire: Agentic UI Design Intelligence
A local design intelligence layer that provides agents with project memory, Figma state, and design system rules for UI development. It facilitates automated design audits, component generation using shadcn/ui, and research-backed specifica…
Multi-model AI Council Deliberation
Initiates a parallel multi-perspective deliberation process by querying Codex, Gemini, and OpenCode. It employs a two-round mechanism to synthesise comprehensive answers from multiple models.
Gemini CLI for Large Codebase Analysis
Initiates an interactive Gemini CLI session, providing a large 1M context window ideal for deep analysis of extensive codebases. It manages both new sessions and follow-up tasks, returning a task ID for immediate result checking.
Adversarial code and security flaw critique
Initiates an adversarial council process where multiple agents collaboratively analyze provided code or answers to identify bugs, security vulnerabilities, and logical flaws. This mode is designed for deep, critical review of complex system…
Query enterprise team screen recording telemetry
This skill enables enterprise administrators to query organizational screenpipe telemetry, accessing data on devices, activity, and transcripts across the entire team. It supports advanced searching and chronological record retrieval using …
Codex CLI for Deep Code Analysis and Reasoning
Initiates an interactive command-line session with OpenAI Codex, enabling deep reasoning, comprehensive code review, and detailed analysis based on user prompts. It manages both new sessions and follow-up continuations.
Screenpipe Health and Diagnostics Agent
This skill enables an agent to monitor Screenpipe's operational status, diagnose audio and vision pipeline issues, and verify database integrity. It provides automated checks for process activity, API responsiveness, and system permissions.
Analyze Screenpipe CLI and Desktop Logs
This skill provides comprehensive log retrieval and analysis capabilities for the Screenpipe ecosystem. It enables developers to diagnose issues by querying both the core CLI backend and the desktop application logs using advanced filtering…
Screenpipe Monorepo Release Automation
Automates version bumping, committing, and triggering GitHub Actions for the Screenpipe monorepo components, including the desktop app, CLI, and MCP. It facilitates monitoring build status and managing the transition from draft to published…
Evidence-backed web research and fact-checking skill
This skill provides verifiable, cited answers by integrating real-time web search results, allowing agents to generate responses with confidence scores and identify source contradictions. It supports both fast, anti-hallucination LLM ground…
Adversarial Code Review for Pipeline Hardening
This skill facilitates the systematic hardening of BFCL training and evaluation pipelines through iterative adversarial review rounds. It employs an external LLM to identify potential bugs, which are then verified against the codebase to en…
Structured planning and assertion generation skill
This skill mandates the creation of a structured test assertions file, guiding automated verification runners. It allows developers to categorize tests by data, pipeline, or agent layers, setting severity gates from warning to immediate abo…
ABA-based Protocol for Precise Agent Execution
This protocol enforces rigorous, observable goal setting and mandates step-by-step execution where every state-changing command must be independently verified against the observable state. It prevents the reinforcement of bad patterns by re…
Automated Pull Request Review and Fix
Systematically analyses open pull requests for CI failures, review comments, and code quality issues, then applies fixes to the relevant branches.
Authoring and Refining Agentic Skills
This skill guides the end-to-end process of designing, improving, and evaluating reusable agent skills. It provides structured methodologies for enhancing trigger quality, auditing skill collections for overlap, and ensuring robust, testabl…
Orchestrates periodic repository quality reviews and escalations
This skill coordinates comprehensive, periodic repository audits, dispatching findings to specialized reviewers for API contracts, documentation, and code quality. It aggregates results, normalizes findings, and escalates outcomes into repo…
Standardized planning workflow management skill
This skill standardizes the planning process by guiding users through selecting appropriate execution modes (spec or no-spec) based on task scope and complexity. It enforces structured workflows, ensuring necessary confirmations and maintai…
CloudBase API Contract Review
Audits CloudBase cloud API wrappers and MCP tools for discrepancies between implementation and official documentation. It verifies parameter casing, request shapes, and interface names to ensure contract correctness.
End-to-end attribution issue triage and repair
This skill manages the full lifecycle of MCP attribution issues, providing an auditable workflow that moves beyond simple status updates. It diagnoses failures using run traces, isolates repairs in dedicated worktrees, and drives the fix th…