Roberto Stagi
Community@rstagi · Milan
Context Engineer @ratel-ai
Agent Skills by Roberto Stagi
Showing 20 vetted skills indexed across 1 GitHub repositories.
rstagidaily-takeaway
Distills today's Claude conversations and memory files into shareable social media insights.
rstagitalk-slides-build
Builds conference talk decks as React apps or reveal.js vector PDFs from slide specs.
rstagimultiphase-plan
Turns a goal into a phased implementation plan registered on a local Loop daemon.
rstagitalk-roast
Fact-checks and red-teams a finished talk, writing a severity-ranked roast report.
rstagitalk-refine
Reconciles cross-artifact coherence across a finished talk's brief, slides, delivery plan, style spec, and built deck.
rstagiupdate-useful-network
Adds and updates non-investor contacts in a Notion CRM database using Gmail and Granola context.
rstagitalk-slides-content
Converts a talk content brief into a slide-by-slide content specification with assertion titles and layout classes.
rstagitalk-delivery
Drafts trigger-word speaker cue cards and a delivery plan from a talk content brief and slide spec.
rstagiloop-execute
Executes published multi-phase plans autonomously with parallel headless runners and merge orchestration.
rstagiops-publish-to-ametyst
Publishes local .claude skills and loops to Ametyst via the runCompound MCP tool.
rstagitalk-slides-style
Generates a slide style spec with typography, palette tokens, and a paste-ready accents.css block.
rstagiskill-creator
Create, evaluate, and iteratively improve Claude skills with benchmarks and description optimization.
rstagilearning-now
Captures session insights into a self-learning log and knowledge index.
rstagifrontend-design
Generates production-grade frontend interfaces with distinctive typography, color, motion, and layout choices.
rstagitalk-content
Generates a conference talk content brief with throughline, audience map, narrative skeleton, evidence, and close.
rstagiloop-pickup
Resumes multi-phase plans in fresh worktrees by recovering context and claiming unblocked lanes.
rstagipr-review
Reviews pull requests by running them in a git worktree and auditing against a fixed checklist.
rstagicodify-from-insights
Triage /insights suggestions and codify approved candidates into skills, hooks, or CLAUDE.md edits.
rstagiloop-handoff
Reconcile worktree progress against a multi-phase plan and repush it to the Loop daemon and Kestral.
rstagitalk
Orchestrates a six-stage pipeline that builds conference talks from inputs to a built React slide deck.
Frequently Asked Questions About Roberto Stagi
FAQPage SchemaWhat tasks can I accomplish with rstagi's skills?▼
You can build conference talks end to end (content brief, slide structure, speaker notes, styling, React or reveal.js deck, adversarial fact-check roast), plan and autonomously execute multi-phase engineering efforts across git worktrees, review pull requests by actually running them, create and benchmark new skills, and log contacts to a Notion network.
Who are these skills designed for?▼
They target engineers and technical speakers using Claude Code or Codex: conference presenters needing decks and delivery coaching, developers running parallel multi-phase work in Conductor worktrees, reviewers auditing PRs, and skill authors who want to create, evaluate, and optimize skills with variance analysis.
How do the talk and loop skills run in practice?▼
Invoke /talk to orchestrate six sub-skills producing numbered artifacts (01-content-brief through 06-roast-report) plus a deck/ directory. For engineering, multiphase-plan registers a plan on the local Loop daemon, loop-execute spawns headless runners per phase, and loop-pickup or loop-handoff resume work across sessions.
What are the prerequisites and dependencies?▼
Skills rely on Claude Code primitives: Bash, Read, Write, Edit, Glob, Grep, WebSearch, WebFetch, and the Skill tool. Loop skills require the Loop daemon and optionally Kestral; update-useful-network needs Notion, Gmail, and Granola access; talk-slides-build scaffolds React + Vite + Framer Motion or reveal.js.
Are these skills open source and what do they cost?▼
The manifest lists them as public skills under Roberto Stagi's GitHub account (rstagi, ratel.sh) with no stated license or pricing in the frontmatter. They are plain skill folders with native frontmatter, publishable verbatim to the Ametyst registry via the ops-publish-to-ametyst skill.