skills/diff-it/SKILL.md
--- name: diff-it description: Open a visual diff for a commit range in the browser. Builds a self-contained, inline-styled HTML page from `git diff <a>..<b>` (no external assets, no CDN, no diff2html dependency) and opens it with `open`. Use when the user wants to eyeball the changes across a set of commits rather than read raw terminal diff output, e.g. "diff the vectorizing work across these 5 commits". Triggers on "diff it", "show me the diff between those commits", "launch a diff viewer", "
npx skillsauth add RonanCodes/ronan-skills skills/diff-itInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Turn a git commit range into a coloured HTML diff and open it in the browser. The point is a quick visual read of "what changed across these commits" without squinting at terminal output or pulling up a PR. The output is one self-contained HTML file: inline CSS only, no CDN, no JS, no diff2html. It renders the same offline, on a plane, or forwarded in an email.
Do NOT fire for:
git diff)./ro:completion-report, the heavier sibling. diff-it is the light,
ad-hoc "just show me this range" tool.One script, scripts/diff-it.sh:
git diff accepts; defaults to HEAD~1..HEAD).git diff <range> and git diff --stat <range>.+ additions green, - deletions red, @@ hunk headers, diff --git
file headers, everything else context).--out).opens it (macOS open, falls back to
xdg-open on Linux). --no-open writes and prints the path only.# Resolve the range from context, then run:
bash "${CLAUDE_PLUGIN_ROOT:-$HOME/Dev/ronan-skills}/skills/diff-it/scripts/diff-it.sh" \
HEAD~5..HEAD --title "vectorizing work"
# Two explicit commits
bash .../diff-it.sh abc123..def456
# Last commit (no range needed)
bash .../diff-it.sh
# A branch comparison, write to a known path, don't pop the browser
bash .../diff-it.sh main..feature --out /tmp/feature-diff.html --no-open
The user usually describes the range in words ("between those commits", "the last five", "since I branched"). Map it to a git range before calling the script:
HEAD~N..HEAD<older-sha>..<newer-sha> (older first)main..HEAD (or origin/main..HEAD)<base>..<head> (or use /ro:completion-report instead)If the range is ambiguous, ask which commits before running. Pass a human title
with --title so the report header reads well (e.g. the feature name, not the
raw range).
Binary files ... differ meta line, the script does not
try to render them.[[completion-report]]: the heavier per-PR / per-story HTML report with CI
status, file stats, risk flags, and per-file rollback. Use that at the tail of
an autonomous run; use diff-it for a quick ad-hoc range.[[visual-diff]] / [[variant-compare]]: image and variant comparison, not
code diffs.testing
--- name: linear-pipeline description: The Fable orchestrator for a single dispatched Linear ticket. Holds almost no context itself; it receives `--issue <ID> --detached`, decides the stage sequence, and fans out a sub-agent per stage, passing forward only each stage's artifact (never re-derived, never inlined into its own context). Step zero, before any planning or stage routing, is a boundary triage against `canon/security-boundary.md` (#199): a match tags Ronan Connolly and stops the run, no
development
--- name: in-your-face description: Capture a chat-only answer into a durable artifact (markdown + HTML, PDF when cheap) and launch it automatically so the user cannot miss it. Use when user says "in your face", "don't let me lose this", "save that answer", "make that durable", or right after answering a substantive side question (a recipe, comparison, how-to, or generated prompt) that would otherwise die with the context. category: workflow argument-hint: [--no-open] [--vault <short>] [hint of
tools
One-shot headless OpenAI Codex CLI calls for background/admin AI tasks — summaries, classification, extraction, admin glue. The default engine for anything that runs AI constantly in the background (daemon-driven, per-event), because it bills the flat ChatGPT subscription instead of Claude usage or per-token API spend, and it keeps working while Claude is rate-limited. NEVER for coding — coding stays Claude. Use when a skill or daemon needs a cheap always-on AI call, when the user says "use codex", "ask codex", "codex as backup", or when building a background summarizer/classifier into a listener or loop. Reads auth from ~/.codex/auth.json (ChatGPT account, no API key).
research
Turn a warranty rejection, repair quote, or RMA email into a cited decision brief — legal read (NL/EU consumer law), is the part user-serviceable, live part and new-unit prices, repair-vs-DIY-vs-new economics, before-you-send-it checklist, deadlines. Use when the user pastes or screenshots a repair quote, warranty rejection, "not covered" email, onderzoekskosten fee, or asks "should I repair or replace this".