Jon Lombardi
Community@jl-cmd · Florida
Automation engineer and full-stack developer. Python, React/TypeScript, Django. Building with AI tools daily.
Agent Skills by Jon Lombardi
Showing 63 vetted skills indexed across 1 GitHub repositories.
jl-cmdparallel-debug
Runs the pr-converge loop across open GitHub PRs using AHK auto-continue pacing in Cursor.
jl-cmdverify
Verifies the claude-dev-env installer lifecycle in a disposable sandboxed home directory.
jl-cmdpr-name-by-capability
Audit GitHub pull requests for driver-word naming violations on reusable capability modules.
jl-cmdpr-plain-language-cleanup
Trims filler words and needless code from every changed file in a pull request.
jl-cmdpr-cleanup
Refines pull requests through parallel audits, focused sizing, and review loops.
jl-cmdpr-shared-extraction
Audits pull requests for misplaced helpers and extracts them into shared_utils modules.
jl-cmdupdate
Fast-forwards a local repository's main branch to a remote's main after confirmation.
jl-cmdsession-tidy
Audits, cleans, and consolidates session logs in an Obsidian vault.
jl-cmdrun-claude-dev-env
Build, install, and smoke-test the claude-dev-env installer against a sandbox home directory.
jl-cmdskill-builder
Orchestrates the lifecycle of building, improving, and polishing Claude Code skills.
jl-cmdpr-small-cl
Scopes or splits pull requests into self-contained reviewable increments.
jl-cmdpr-refinement
Coordinate pull-request audits and publish dependency-ordered GitHub pull-request stacks.
jl-cmdhitl
Enforces human-in-the-loop scope reporting and bounds worker orchestration.
jl-cmdautoconverge
Drives a draft GitHub PR to convergence through autonomous review and fix rounds.
jl-cmdsession-log
Generates HTML session reports with frontmatter metadata and publishes them to an Obsidian vault.
jl-cmdprototype
Builds proof-of-concept code in a hookless git worktree sandbox, then promotes it through clean-room review.
jl-cmdreview-tier
Classify pull-request change surfaces into T1, T2, or T3 review tiers from git evidence.
jl-cmdrebase
Rebases git branches onto their base ref with verification gates before force-pushing.
jl-cmdpr-title-description
Writes plain-language pull request titles and descriptions from full diffs.
jl-cmdsource-command-sr-loop
Runs a converging cleanup loop repeating simplify and code-review passes until a PR diff is clean.
jl-cmdpr-loop-cloud-transport
Routes PR-loop skill operations through GitHub MCP tools when gh CLI is unavailable.
jl-cmdbugteam
Automates multi-loop pull request audit and fix cycles until convergence on GitHub.
jl-cmdrelease-notes-html
Generates a self-contained HTML release-notes page from session evidence and repository state.
jl-cmdpr-fix-protocol
Applies reviewer findings to pull requests as verified fixes and resolves review threads.
Frequently Asked Questions About Jon Lombardi
FAQPage SchemaWhat tasks can I automate using jl-cmd's skills?▼
You can run PR convergence loops (pr-converge, autoconverge, bugteam) that cycle Bugbot, Copilot, and Codex reviews with fixes until clean; audit PRs for naming and extraction issues; rebase and split branches; file GitHub issues; and journal sessions to an Obsidian vault.
Who are these skills built for?▼
They target automation engineers and full-stack developers working in Claude Code or Cursor who manage GitHub pull requests at high volume. The author's stack is Python, React/TypeScript, and Django, and several skills assume that environment, such as the pytest/Playwright test-runner.
How do I install and run these skills?▼
Skills install through the claude-dev-env installer, which the run-claude-dev-env skill builds, smoke-tests in a sandbox, and verifies in a disposable home. Once installed into ~/.claude, each skill triggers on its slash command or natural-language phrase listed in its description.
Are jl-cmd's skills open source and free to use?▼
The skills are published as public repositories under the jl-cmd GitHub account with no stated licensing fee or paid tier. They run inside your own Claude Code or Cursor sessions, so usage costs are limited to your own model and reviewer quotas, such as the once-per-run Copilot pre-check.
What prerequisites do these skills require?▼
Most PR-loop skills require a GitHub repository, the gh CLI with auth (or the pr-loop-cloud-transport fallback), and Claude Code with hook support. Specific skills need extras: Obsidian for session journaling, Windows with PowerShell for scheduled tasks, Everything (es.exe) for file search, and Django/Playwright for the test-runner.