Anthony James Padavano
Community@4444j99 · New York City
Builder of production systems — data, infra, AI, deploy, end-to-end. Python · TypeScript · React · FastAPI · Docker.
Agent Skills by Anthony James Padavano
Showing 9 vetted skills indexed across 2 GitHub repositories.
4444J99agy-conductor
Connects Antigravity agents to Limen's peer-conductor protocol for bounded task execution.
4444J99decorum-voice-judge
Score public prose on tone, brand voice, clarity, and credibility with content-pinned verdicts.
4444J99closeout
Drive tasks to verified closeout with zero open items and relay handoffs.
4444J99experience-judge
Scores visual quality of captured website screenshots and records verdicts in a YAML judgment register.
4444J99lesson-harvest
Extracts lessons from chat session transcripts and routes them to durable owners.
4444J99verify
Verifies limen repository changes by driving the correct runtime surfaces and avoiding measured verification traps.
4444J99vendor-insights
Generates faceted usage reports and HTML narratives for non-Claude agent session stores.
4444J99ship
Drives a change from implementation through scoped gates, PR merge, and live-artifact verification.
4444J99script-analysis-dramaturgical
Analyze screenplays to generate eight dramaturgical coverage documents.
Frequently Asked Questions About Anthony James Padavano
FAQPage SchemaWhat tasks can I accomplish with Anthony James Padavano's skills?▼
You can orchestrate bounded work through the Limen peer-conductor protocol, ship changes to live-verified merges, close out tasks to zero dangling items, judge public prose voice and visual page quality, harvest chat sessions into durable lessons, generate cross-vendor usage reports, and produce eight dramaturgical coverage documents from screenplays.
Who are these skills designed for?▼
These skills target engineers and operators running multi-agent estates (Claude, Codex, Copilot, Opencode, Antigravity/Agy, Gemini) who need release discipline, session-to-lesson processing, and quality judging of public surfaces, plus screenplay analysts needing dramaturgical coverage generation.
How do the ship, verify, and closeout skills work together at runtime?▼
Read verify first to learn the repo's drive paths and gotchas, then ship drives one change through scoped gates, PR, required-checks merge, and live-artifact proof. Closeout finishes by verifying ground truth read-only, closing every gap per owner, reaching an idempotent fixed point, and producing a relay handoff.
What prerequisites do the judging and harvesting skills require?▼
Each model-in-the-loop skill pairs with an existing mechanical script: decorum-voice-judge needs scripts/decorum-keeper.py, experience-judge needs scripts/experience-audit.py, lesson-harvest needs scripts/lesson-harvest.py with its backlog sensor, and vendor-insights needs scripts/vendor-insights.py for session indexing and HTML rendering.
Can I generate usage insights for non-Claude agents on my host?▼
Yes. The vendor-insights skill produces a Claude-/insights-style faceted usage report for any other agent estate on the host, including codex, copilot, opencode, antigravity/agy, and gemini, answering questions like 'what has this lane been doing' with bounded transcript reads and HTML rendering.