9,963 tagged LLM, measured the same way as everything else here.
Browse within: agent 496chatgpt 228agents 218cli 169ml 133javascript 124bun 123agentic-ai 122openai 119agent-framework 116deep-learning 107foundation-models 98agentskills 90cortex 90
williamcorrea23/androidbuilder-mcp
Skill Codex
Automate and audit MIT App Inventor, AndroidBuilder, Kodular, Niotron, and AppyBuilder projects through the androidbuilder-mcp MCP server. Use whenever the user wants to read or change anything in one of these browser IDEs — screens, component trees, component properties, Blockly blocks (inspect/export/add/clear XML)…
Skill Claude CodeCodex
Distill a git repo's full history into durable engineering memory under .repo-memory/ (timeline, decisions, landmines, architecture). Use when the user wants historical/"why is this like this" context for a codebase, to onboard an agent onto a repo's past, or asks to run repo-history / analyze git history.
Skill Claude Code
Plan a repository task for execution by a small local coding model, then hand the plan to the local executor over MCP. Inspect the codebase, decompose the work into self-contained sequential prompts, define deterministic eval gates, validate the manifest, and submit it for execution. Never implement the task.
Skill Claude CodeCodex
A data-quality repair process in which one role proposes database fixes and another independently checks them on a copy before changes are saved.
Skill Claude CodeCodex
A code-review workflow written in Chinese that checks software changes for correctness, security, performance, and maintainability. It also defines a report format for critical issues and suggestions.
Skill Claude CodeCodex
Retrieve scoped project context and record evidence-backed handoffs through the Context Forge MCP server.
AvalancheAI-labs/agent-wiki-kit
Skill Claude CodeCodex
Compile a raw source (document, URL content, meeting notes, research output) into a wikikit agent wiki — merge-don't-overwrite, log contradictions, maintain cross-references, then lint. Use whenever the user says "add this to the wiki", "ingest this", or drops new source material for a wiki maintained with…
Skill Claude Code
Use when verifying that a tool or skill was actually used, checking tool installation, or generating verification receipts. Triggers on 'verify', 'truth', 'proof', 'did it actually use', 'check tool', 'tooloftruth', '/truth'. Also triggers when the agent is about to claim a tool was used — run verification first.
Skill Claude CodeCodex
Converts CSV exports to Parquet and validates them against a declared schema, reporting type mismatches per column. Use when the user asks to convert, compact, or validate tabular data files before loading them into a warehouse.
Skill Claude CodeCodex
How this household runs its Sunday weekly review.
Skill Claude CodeCodex
Converts PDF documents, Word files (.docx), CSV spreadsheets, and YouTube video URLs into clean, token-efficient Markdown with smart table and heading structure detection.
Skill Claude Code needs its repo
Run ONE iteration of the Kumihimo build loop — take the next item off the queue, ground it against PLAN.md, build it, verify it, document it, commit it. Use when Thomas invokes /kumihimo-iteration (often wrapped in /loop for unattended runs) or asks for another build pass on Kumihimo. Covers the queue, verification…
Skill Claude CodeCodex
Control Spotify and build LLM-reasoned playlists from the terminal — play/pause/skip/volume, recommend music in the vibe of a playlist, generate mood playlists, and summarize taste. Use when the user wants to control Spotify, get music recommendations, or build/curate playlists. Requires the spotify-llm-mcp MCP server…
yourlastnamesoundslikeatypeofpasta/acumatica-mcp-server
Skill Claude CodeCodex
Run an accounts payable health check. Use when the user asks "what do we owe?", "are any vendor bills overdue?", "show AP aging", "are there receipts we haven't been billed for?", "any old open POs?", or wants a full payables summary.
Skill Claude CodeCodex
Extract text from PDFs/scans (pymupdf, marker-pdf).
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: