By MarjovanLier
Skills for maintaining a Karpathy-style LLM wiki: ingest knowledge, look it up, run checkpoints, and lint the wiki.
Run a wiki and memory checkpoint at the end of a chat thread or before context compaction. Use when the user invokes wiki-checkpoint, asks to save session knowledge, asks for a wiki or memory sweep, or wants durable findings, retrospective lessons, and routing decisions captured from the current conversation. Honour focus text and skip flags such as skip wiki, skip memory, and skip retrospective.
Ingest one explicitly identified input into the wiki as a single transaction. A URL, file, pasted text, command output, stated fact, or an investigation result from the current session. Use when the user asks to add a specific input or finding to the wiki, says "wiki this" or "ingest this into the wiki", or asks to record a conclusion, decision, or finding permanently. Generic "ingest this" or "file this" without wiki context is not this skill. Not for behavioural rules or memory updates, not for end-of-thread or pre-compaction knowledge sweeps (use wiki-checkpoint), and not for wiki health checks (use wiki-lint).
Lint and re-index a Karpathy-style LLM wiki (second brain). Use this whenever the user asks to "lint the wiki", "wiki lint", "check wiki health", "index the wiki pages", "refresh the wiki search index", or anything about validating wiki structure or making wiki pages searchable. The wiki is sometimes a project-local folder and there is always a global one at ~/wiki. This skill detects which wiki applies, runs the correct lint script (which also refreshes the qmd search index in one pass), and resolves flagged issues. Do NOT use it for ordinary source-code linting (eslint, phpstan, markdownlint of non-wiki files) or for the full pre-compaction knowledge sweep (that is /wiki-checkpoint).
Read-only lookup of existing knowledge in the user's Karpathy-style wikis, qmd-backed where a collection exists. Use whenever the user asks what the wiki knows or says about something, whether a topic was researched or decided before, says "look this up in the wiki", "check the wiki", "have we covered this", or asks "what do we know about X" in the sense of accumulated past project, personal, case, setup, or cross-project knowledge. Do not use for ordinary questions about the current codebase, repository documentation, live system state, or general knowledge. Not for adding knowledge (wiki-ingest), end-of-thread sweeps (wiki-checkpoint), or wiki health, sync, and index maintenance (wiki-lint).
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.
A collection of plugins for Claude Code CLI.
| Plugin | Description |
|---|---|
| git-conventions | Git workflow skills: Conventional Commit messages and MR/PR contribution descriptions |
| wiki-toolkit | Skills for maintaining a Karpathy-style LLM wiki: ingest, lookup, checkpoint, lint |
| intelephense-lsp | PHP language server for code intelligence, completion, and diagnostics |
| php-lspx | PHP language server: intelephense and phpactor multiplexed via lspx |
| php-code-simplifier | Agent that simplifies and refines PHP code for clarity and maintainability |
Add this repository as a marketplace to browse and install plugins:
# From GitHub
/plugin marketplace add MarjovanLier/claude-plugins
# Or from local clone
/plugin marketplace add /path/to/claude-plugins
Then install plugins:
/plugin install plugin-name@marjo-claude-plugins
Load a specific plugin directly:
claude --plugin-dir /path/to/plugins/plugin-name
Each plugin follows this structure:
plugin-name/
├── .claude-plugin/
│ └── plugin.json # Required: Plugin manifest
├── README.md # Plugin documentation
├── commands/ # Slash commands (/command-name)
│ └── command-name.md
├── skills/ # Model-invoked capabilities
│ └── skill-name/
│ └── SKILL.md
├── agents/ # Autonomous workers
│ └── agent-name.md
└── hooks/ # Event-driven automation
└── hooks.json
plugins/.claude-plugin/plugin.json with plugin metadatamarketplace.json in the root .claude-plugin/ directoryclaude --plugin-dir /path/to/your/plugin# Load a specific plugin
claude --plugin-dir ./plugins/php-code-simplifier
# Verify commands are registered
/help
# Test agent availability
# The agent should appear in Task tool options
MIT
npx claudepluginhub marjovanlier/claude-plugins --plugin wiki-toolkitGit workflow skills: Conventional Commit messages and merge request contribution descriptions.
PHP language server: intelephense and phpactor multiplexed via lspx for completion, diagnostics, and refactoring.
Agent that simplifies and refines PHP code for clarity, consistency, and maintainability while preserving functionality
Fun thinking-style skills and an agent: first-principles breakdowns and What Would Elon Do analysis.
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
v9.54.1 — Reliability wave: tangle contextual review correction loop with hard round ceiling, progress-supervised review rounds (per-agent stall watch, descendant-tree kills), council diversity and agy pin fixes, marketplace generator source-of-truth fix, provider troubleshooting runbook and cost-expectations docs. Run /octo:setup.
Build and maintain an LLM-curated personal knowledge base with default local FastEmbed + sqlite-vec search, BM25 fusion, structured evidence, incremental caching, and a compiled graph layer.
Connect to Atlassian products including Jira and Confluence. Search and create issues, access documentation, manage sprints, and integrate your development workflow with Atlassian's collaboration tools.
AI-powered wiki generator for code repositories. Generates comprehensive, Mermaid-rich documentation with dark-mode VitePress sites, onboarding guides, deep research, and source citations. Inspired by OpenDeepWiki and deepwiki-open.
Remove signs of AI-generated writing from text, making it sound more natural and human. Based on Wikipedia's "Signs of AI writing" guide.