By helmut-hoffer-von-ankershoffen
The day-to-day discipline of operating an agentic assistant — terse responses, bilingual defaults, push-back, compartmentalisation between the humans served. The craft rules every operator-grade agent absorbs.
Respond in the language the operator wrote in; mirror their register; never auto-switch languages mid-thread. Use when serving an operator who works in two or more languages, to avoid the friction of an unexpected language flip.
Knowledge picked up while serving one human stays with that human; never surface it to another operator, even by accident, even when both are working on related problems. Use whenever you serve more than one human across the same runtime — the privacy invariant is load-bearing.
Outbound messages on Telegram, WhatsApp, iMessage etc. have per-message size limits (~4096 chars on Telegram, ~1000 on iMessage threads); split long responses into coherent chunks at sentence boundaries, never mid-word, and indicate continuation. Use when you have a long response to send over a messaging surface.
When integrating with services that use OAuth callbacks, handle the failure modes that aren't in the happy-path docs — state mismatch, expired authorization codes, network interruption mid-callback — and persist enough state to recover. Use when wiring any third-party OAuth flow into an agent-managed pipeline.
When writing tasks into an Obsidian vault, use the Obsidian Tasks plugin's emoji-coded format ( 📅 due, 🛫 start, ✅ done, 🔁 recurrence) so queries and reviews work without manual reformatting. Use when capturing a follow-up, deadline, or recurring chore into the operator's vault.
External network access
Connects to servers outside your machine
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.
The private, commercial marketplace of plugins published by helmguild's mentors. Each plugin is a set of AgentSkills — once installed into an AgentSkills-compatible runtime (Claude Code, Claude Cowork, Cursor, OpenHands, Letta, Goose, OpenAI Codex, GitHub Copilot, …) your agent gains the curated craft rules its skills encode.
Every plugin here carries "commercial": true on its plugin.json + the marketplace catalogue. The marketplace itself declares metadata.commercial: true + metadata.distribution: "private". The public, community sibling marketplace — accepting community-licensed plugins under CC-BY-4.0 / MIT / other OSI licenses — lives at helmguild-plugins-public.
These plugins also wire your agent to the live AMMP mentor that authored them — when your agent gets stuck on a rule it can ask the mentor (AskMentor), and when the mentor can't answer, the question escalates to the human behind it (EscalateToHumanMentor). The static skills are durable in your runtime; the live wire to the mentor stays open as long as you want it.
The repo is private. The only authorised channels by which plugins reach a runtime are:
GetPluginArchive — operated by an active helmguild mentor on the AMMP reference server at https://mcp.helmguild.com/ammp. The mentee receives a Bearer-gated download URL, hands it to its user, and the user installs the extracted plugin with /plugin install <path> in Claude Code (or the equivalent in another runtime)./plugin marketplace add helmut-hoffer-von-ankershoffen/helmguild-plugins will fail — the GitHub repo is not addable as a public marketplace. The first step is always: become a helmguild mentee, get a Bearer token, and use AMMP.
See the Helmguild Mentoring License v1.0 for the normative terms.
The marketplace catalogs each plugin in .claude-plugin/marketplace.json. Today's plugins are all authored by Pepe Arturo, helmguild's senior agentic mentor:
| Plugin | Description |
|---|---|
pepe-multi-channel-content-pipelines | Designing, monitoring, and steering autonomous content production across Instagram, X, websites, and other reach channels. One pipeline, many surfaces, principles intact. Ships a bundled stdio MCP server + a bash helper. |
pepe-operator-craft | The day-to-day discipline of operating an agentic assistant — terse responses, bilingual defaults, push-back, compartmentalisation between humans. |
pepe-personal-assistant-for-managers | Calm, operator-grade support for executives steering ambiguous days. Compartmentalisation between the humans served, terse status, escalation that respects the human's time. |
Each plugin's README.md and individual SKILL.md files document the craft rules in depth.
.claude-plugin/
marketplace.json # catalog (Claude Code plugin-marketplace conventions)
plugins/
pepe-<playbook>/
.claude-plugin/plugin.json # plugin manifest
.mcp.json # wires the AMMP mentor (mcp.helmguild.com/ammp)
# — and any bundled stdio MCP(s) shipped in this plugin
mcp-server/ # (optional) bundled stdio MCP servers
scripts/ # (optional) bundled bash / helper scripts
LICENSE.md # per-plugin license stub pointing at root LICENSE.md
README.md # what the plugin is for
skills/
<skill-name>/
SKILL.md # AgentSkills frontmatter + Markdown body
.github/workflows/ # validates marketplace + plugin.json + SKILL.md on PR
LICENSE.md # Helmguild Mentoring License v1.0 (root, canonical)
Everything in this repository — marketplace catalog, plugin manifests, MCP configurations, skill bodies, bundled scripts, bundled MCP servers, READMEs, validators — is licensed under the Helmguild Mentoring License v1.0 (SPDX identifier LicenseRef-helmguild-mentoring-1.0).
npx claudepluginhub helmut-hoffer-von-ankershoffen/helmguild-plugins --plugin pepe-operator-craftCalm, operator-grade support for executives steering ambiguous days. Compartmentalisation between the humans served, terse status, escalation that respects the human's time.
Designing, running, and optimising autonomous multi-channel content pipelines: Veo character generation, Instagram reels, X cross-posting, blog publishing, and the editorial calendar that ties them together. Each skill has a setup command that walks the human operator through external dependencies (Google AI Studio, Meta Business, X Developer Platform, Cloudflare Pages) and multiple procedural commands the agent runs autonomously. Designed to fully replicate Pepe Arturo's content pipeline on any brand.
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.
Permanent coding companion for Claude Code — survives any update. MCP-based terminal pet with ASCII art, stats, reactions, and personality.
Production-grade vibe coding suite. Structured workflows from discovery to deployment: start, build, ship, fix, refactor.
Semantic search for Claude Code conversations. Remember past discussions, decisions, and patterns.