By yugabyte
Meko MCP tools for memory, conversations, knowledge base, and datapacks.
Behavioral guide for Meko MCP tools on desktop Claude. READ IT BEFORE calling any meko tool — it sets the required agent_id and workflow; calling a meko tool without it produces wrong-bucket writes. Use whenever the user mentions memory, Meko, datapacks, knowledge base, or conversations, or asks to remember/save/store anything. ALSO trigger (even with no tool or Meko keyword) when: (a) the user states ANY durable fact about themselves — name, role, team, tools, stack, preferences, conventions — INCLUDING facts dropped inside a task/technical question ("I use pnpm", "I'm evaluating X"); or (b) the user asks a RECALL question — "what do you know about me", "who am I", "what was the last thing I told you" — then search Meko (memory_search) first and never claim memory is unavailable without checking. On Desktop the agent_id is ALWAYS "claude_desktop" (or "meko_agent" for cross-project identity) — NEVER "claude_code".
Behavioral guide for AI agents using Meko MCP tools. Triggers when calling MCP tools, storing memories, persisting knowledge, managing datapacks, or searching shared knowledge bases through Meko.
Lists the user's Meko datapacks, shows authorization, supports search and pagination, and pins one as the active datapack on Claude Desktop. Triggers when the user says things like "switch datapack", "use the Acme datapack", "list my datapacks", "which datapack am I on", "pin Acme", or any request to change or inspect the active datapack. Persistence is via memory_add (Desktop has no SessionStart hook), so the pin survives across sessions until the user clears it.
Lists the user's Meko datapacks, shows authorization, supports search and pagination, and pins one as the active datapack for the current project. Triggers when the user says things like "switch datapack", "use the Acme datapack", "list my datapacks", "which datapack am I on", "pin Acme", or any request to change or inspect the active datapack. Also trigger before the first Meko MCP write of a session if no datapack pin exists and the user has more than one datapack.
External network access
Connects to servers outside your machine
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Off-the-shelf agent skills for Meko — the agent-native data layer for multi-agent systems.
Meko works without skills. Once you connect the MCP server to any AI agent, the agent has access to tools for memory, conversation history, knowledge base, and datapack management. See the full list of MCP tools available.
You tell it what to save and when:
"Save to memory that I'm a backend engineer and we use Go."
"Search my memories for what we discussed about the auth migration."
That's on-demand mode — Meko is a tool your agent calls when you ask it to. Many users start here and it works fine.
Skills change what happens next. A skill is a markdown instruction file (SKILL.md) that teaches your agent to use Meko without being asked:
| Without skill | With skill installed |
|---|---|
| You say "save this to memory" | Agent calls memory_add the moment you mention your name, role, or preferences |
| You guess at tool parameters | Agent uses correct scope, agent_id, and datapack_id routing on the first try |
| You ask the same questions across sessions | Agent calls memory_search at session start and greets you with context |
| Conversations vanish when the window closes | Agent preserves exchanges via conversation_create + conversation_add_message |
Each layer builds on the previous:
| Layer | What you get | What you install |
|---|---|---|
| MCP only | Tools available on demand — you tell the agent when to save | MCP server connection (setup guide) |
| MCP + Skill | Agent proactively stores memories, classifies info, uses correct parameters | + a skill from this repo |
| MCP + Skill + Hooks | Full automatic conversation capture — every exchange persisted without prompting | + hooks (Claude Code, via the installer) |
This repository has three public-facing surfaces with different consumers:
| Directory | Consumer | Purpose |
|---|---|---|
skills/ | Humans and harnesses that copy skills directly (Cursor, Codex, Copilot, Claude Desktop, claude.ai, etc.) | Browsable community skill source at stable paths |
plugins/meko-agent-skills/ | Claude Code /plugin and Connector Directory packaging | Self-contained plugin bundle with its own generated copy of skills/, hooks, hook handlers, and .mcp.json |
scripts/ | CI and publish verification | Token-free checks for generated public output, prompt-injection lint, and plugin readiness |
.claude-plugin/
└── marketplace.json # Marketplace catalog; metadata.pluginRoot points at ./plugins
plugins/
└── meko-agent-skills/
├── .claude-plugin/plugin.json
├── .mcp.json # Public Meko MCP connector config
├── skills/ # Generated copy; plugin must not ../ into root skills
├── hooks/
└── hooks-handlers/
skills/
├── meko-mcp-tools/ # Coding agents: Claude Code, Cursor, Codex, VS Code
└── meko-mcp-tools-desktop/ # Claude Desktop, claude.ai (no session hooks)
scripts/
The duplication between skills/ and plugins/meko-agent-skills/skills/ is intentional. Anthropic plugins are copied into a plugin cache and cannot reference files outside their plugin directory with paths like ../shared-utils, so the plugin must carry its own skill copy.
Both skills cover 20 MCP tools across memory, conversation, knowledge base, and datapack management. The difference is how they handle session lifecycle:
| Skill | Best for | Key difference |
|---|---|---|
meko-mcp-tools | Claude Code, Cursor, Codex CLI, GitHub Copilot (VS Code / CLI) | Designed for coding agents — includes hook-based conversation capture and subagent coordination |
meko-mcp-tools-desktop | Claude Desktop, claude.ai | Designed for chat-first clients — more aggressive proactive-memory rules, no hook dependency |
You need a Meko account and an MCP server connection before installing a skill. If you haven't set that up yet:
npx claudepluginhub yugabyte/meko-skills --plugin meko-agent-skillsUse when writing or generating Terraform for the yugabyte/yba provider — installing YugabyteDB Anywhere (YBA) on a VM, registering the first customer, and creating AWS/GCP/Azure or on-prem cloud providers, storage configs, universes, backups, schedules, and restores. Also covers the upstream cloud IAM (IAM user / service account / service principal) that grants YBA node-deployment permissions and the CLI commands to discover VPC/subnet/security-group network info. Triggers on yba_installer, yba_customer_resource, yba_aws_provider/yba_gcp_provider/yba_azure_provider, yba_universe, yba_backup, source = "yugabyte/yba", or any mention of the YBA Terraform provider. Does NOT cover the YBA REST API directly (use yba-api) or YugabyteDB Aeon.
Evidence-gated AI coding workflow: scan → analyze → plan → TDD → execute → fix → verify → review, powered by Codebase Memory MCP >= 0.9.0 with optional Serena LSP intelligence. Includes blast-radius planning, test/cycle gates, independent review, and Windows Git Bash hook auto-resolution.
A growing collection of Claude-compatible academic workflow bundles. Covers scientific figures, manuscript writing and polishing, reviewer assessment, citation retrieval, data availability, paper reading, literature search, response letters, paper-to-PPTX conversion, and evidence-grounded Chinese invention patent drafting. Rules are organized as reusable skill folders with explicit workflows and quality checks.
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Harness-native ECC plugin for engineering teams - 67 agents, 279 skills, 94 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer
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.