Scaffold a local knowledge management system with VitePress portal, YAML microdatabases, and configurable entity types (ADRs, notes, tasks, discussions). Capture and structure decisions, team discussions, meeting transcripts, and notes into Markdown with YAML frontmatter, then search, review, and refine the knowledge base via semantic and full-text queries.
Use this agent when the user asks to "refine secondbrain", "improve knowledge base", "clean up documentation", "background refinement", or mentions wanting autonomous improvement of their secondbrain with notification support. Examples: <example> Context: User has been working on a project and wants to improve their documentation user: "Can you help refine my secondbrain in the background?" assistant: "I'll launch the secondbrain-refine agent to analyze and improve your knowledge base autonomously." <commentary> The agent runs in background mode, analyzing content quality and proposing improvements with notifications when user confirmation is needed. </commentary> </example> <example> Context: After a sprint, user wants to clean up stale content user: "Review and clean up my knowledge base, notify me on my phone when you need input" assistant: "I'll start the refinement agent with mobile notifications enabled via ntfy.sh." <commentary> Agent uses ntfy.sh for mobile push notifications when awaiting user decisions. </commentary> </example>
Use this agent when the user asks to "review ADR", "review document", "check note quality", "validate discussion", or mentions wanting feedback on a specific secondbrain document before finalizing. Examples: <example> Context: User just created an ADR and wants feedback before proposing it user: "Can you review ADR-0015 before I mark it as proposed?" assistant: "I'll review ADR-0015 for completeness, clarity, and alignment with best practices." <commentary> Agent provides structured feedback on the ADR covering all required sections. </commentary> </example> <example> Context: User wants to improve a note before sharing with the team user: "Review my kubernetes-deployment note for the team" assistant: "I'll analyze the note for clarity, technical accuracy, and actionable content." <commentary> Agent reviews the note focusing on team consumption rather than personal use. </commentary> </example>
This skill should be used when the user asks to "create ADR", "document decision", "architecture decision", "add decision record", or mentions needing to record a technical decision with status tracking and category organization.
This skill should be used when the user asks to "document discussion", "save meeting notes", "record conversation", "capture meeting", or mentions needing to document team discussions with participants, decisions, and action items.
This skill should be used when the user asks to "add entity", "create custom entity", "new entity type", "define entity", or mentions wanting to add a new trackable data type to their secondbrain project beyond the predefined ADR, Note, Task, and Discussion types.
This skill should be used when the user asks for "freshness check", "freshness report", "what needs attention", "stale items", "review backlog", or mentions wanting to see what secondbrain content needs review or updating.
This skill should be used when the user asks to "create a secondbrain", "scaffold knowledge base", "init documentation portal", "set up second brain", "create TDA project", or mentions wanting to build a personal knowledge management system with VitePress and microdatabases.
Modifies files
Hook triggers on file write and edit operations
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Personal plugin marketplace for Claude Code — skills, hooks, and agents for workflow automation and development toolkits.
Marketplace ID: bershadsky-claude-tools
# Add marketplace
/plugin marketplace add sergio-bershadsky/ai
# Install a plugin (e.g., git)
/plugin install git@bershadsky-claude-tools
| Plugin | Description | Skills | Hooks |
|---|---|---|---|
| git | Git workflow automation: conventional commits, auto-staging, uncommitted changes protection | /commit, /version | auto-stage, pre-stop-commit |
| settings-sync | Sync Claude settings across ephemeral VMs via Git backup | — | backup-settings |
| Plugin | Description | Skills | Agents |
|---|---|---|---|
| django-dev | Opinionated Django toolkit with Ninja API, Unfold admin, pytest, Dynaconf | 4 skills | django-review |
| frappe-dev | Frappe Framework v15 toolkit with multi-layer architecture | 5 skills | — |
| Plugin | Description | Skills | Hooks |
|---|---|---|---|
| secondbrain | Knowledge base scaffolding with microdatabases, VitePress portal, semantic search, review stamps, meeting transcription | 11 skills | freshness, sidebar, search-index, session-context, transcription-check |
| replit-prompts | Optimized prompts, PRDs, and task plans for Replit Agent | 3 skills | — |
| Plugin | Description |
|---|---|
| marketplace-validator | Validates marketplace.json schema after every edit |
Knowledge base scaffolding with YAML microdatabases, configurable entity types (ADRs, Notes, Tasks, Discussions, custom), dual semantic search (qmd for CLI, Orama for browser), review stamps with staleness badges, and meeting transcription. 11 skills, 2 agents, 5 automation hooks. See the full documentation.
/plugin install secondbrain@bershadsky-claude-tools
Conventional commits with branch safety checks and ticket detection (GitHub, GitLab, Jira, Linear). Semantic version bumping with git tags. Auto-stages files after edits. Blocks session exit if uncommitted changes exist.
/plugin install git@bershadsky-claude-tools
Skills: /commit, /version | Hooks: auto-stage, pre-stop-commit
Backs up ~/.claude/ to the repo before session ends, excluding sensitive files. Bootstrap script restores settings on new VMs (Codespaces, Gitpod, cloud).
/plugin install settings-sync@bershadsky-claude-tools
Bootstrap: curl -fsSL https://raw.githubusercontent.com/sergio-bershadsky/ai/main/plugins/settings-sync/scripts/bootstrap.sh | bash
Opinionated Django patterns: 1-file-per-model, UUID PKs, soft delete, Dynaconf config, Django Ninja APIs (1-endpoint-per-file, Pydantic schemas), Unfold admin with HTMX, pytest + factory_boy testing. Code review agent for convention compliance.
/plugin install django-dev@bershadsky-claude-tools
Skills: django-dev, django-dev-ninja, django-dev-unfold, django-dev-test | Agent: django-review
Frappe Framework v15 with multi-layer architecture (Controller → Service → Repository → DB). DocType scaffolding, REST API endpoints, service layer patterns, app bootstrapping, pytest integration tests.
/plugin install frappe-dev@bershadsky-claude-tools
Skills: frappe-doctype, frappe-api, frappe-service, frappe-app, frappe-test
Structured prompts for Replit Agent based on Replit's 10 official guidelines. Generates optimized prompts, PRDs, and phased development plans that minimize iteration.
/plugin install replit-prompts@bershadsky-claude-tools
Skills: replit-prompt, replit-prd, replit-plan
Internal. Validates marketplace.json schema after every Write/Edit. Blocks the session if validation fails.
Run documentation site locally:
npm install
npm run docs:dev
Visit: http://localhost:5173
Public domain (Unlicense)
Professional Frappe Framework v15 development toolkit with multi-layer architecture patterns, DocType scaffolding, REST API v2 development, and comprehensive testing
Opinionated Django development toolkit with Ninja API, Unfold admin, pytest, and Dynaconf patterns
Generate optimized prompts, PRDs, and task plans for Replit Agent that maximize AI understanding and minimize iterations
Git workflow automation: conventional commits with branch protection, ticket tracking, and semantic correlation
Semantic version bumping for plugins with git tags and commits
npx claudepluginhub sergio-bershadsky/ai --plugin secondbrainGenerate optimized prompts, PRDs, and task plans for Replit Agent that maximize AI understanding and minimize iterations
AI-powered knowledge base management - Capture conversation learnings, maintain topic-specific KB files, Obsidian-compatible knowledge graph, structured dynamic context loading, and institutional knowledge in CLAUDE.md
Standalone utilities for codebase exploration, web research, git operations, and code maintenance
Persistent memory for Claude Code sessions using Honcho
Plugin for effective agentic development
Knowledge base skills for Claude Code — capture, search, and synthesize project knowledge