Revealui ops

Skill RevealUIStudio/revskills/skills/revealui-ops

Thin shim onto the operational-workflow-layer runner (workflow-run.js). Lists registered workflows (cleanup-session, prepare-for-exit) or runs one by name, passing --dry-run/--fix/--yes straight through. Exposed as /ops. Delegates entirely to the runner — never reimplements workflow execution or safety classification (auto/report-first/gated/owner-only).From its SKILL.md

Install

npx -y skills add RevealUIStudio/revskills --skill revealui-ops

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its file declares

Copied from the file, not written here

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

2.7 KB, as published. Nobody here has run it

Thin CLI shim onto the .jv operational-workflow-layer runner (workflow-run.js, design contract at docs/gap-specs/GAP-314-operational-workflow-layer-design.md §4) — exposed as /ops. This skill delegates entirely to the runner; it does not reimplement workflow execution, step logic, or safety classification.

Load helpers:

. "$HOME/revfleet/revskills/scripts/lib/session-state.sh"

`/ops` or `/ops list` — discovery

node "$JV_REPO/scripts/workflow-run.js" --list

Prints every registered workflow (name, title, and its declared safety — always the max severity across its steps) from $JV_REPO/workflows/*.yml.

`/ops <name> [flags]` — run a workflow

node "$JV_REPO/scripts/workflow-run.js" <name> [--dry-run] [--fix] [--yes]

Pass --dry-run, --fix, --yes straight through — the runner defines what each does; this skill does not interpret them. --dry-run prints the plan and executes nothing.

Safety classes (runner-enforced, not this skill)

Class	Runner behavior
`auto`	Idempotent/read-only/report-only. Runs unattended.
`report-first`	Writes reviewable files (never destructive). Runs and prints; never auto-commits.
`gated`	Destructive or state-changing. Needs `--fix` or `--yes` for that run, or the runner prints `STOPPED-GATED` and halts the chain there.
`owner-only`	Disposition-shaped (per `~/.claude/rules/disposition-actions.md`) — merges, gate-clearing labels, remote branch deletes, repo settings. NEVER executed by the runner under any flag; it prints the exact command for the owner and marks `SKIPPED-OWNER`.

Do not

Do NOT invoke a workflow's underlying step commands directly through this skill — always go through workflow-run.js so safety classification applies. The individual commands remain directly invocable outside this skill (that's Tier 2, the escape hatch); this skill is Tier 3, the named-chain path.
Do NOT pass --fix/--yes without telling the user which gated step(s) it will unlock.
Do NOT attempt to run an owner-only step's printed command through another tool on the runner's behalf — hand it to the owner verbatim, as printed.