webview

A macOS webview interface that opens a native window for collecting structured information or reviewing content with a person.

In plain words
What is it for?
Use it when an agent needs a person to fill in related fields, choose from many options, approve a risky action, inspect a diff, or comment on a Markdown document.
Why use it?
It makes multi-field input, several-option choices, approvals, previews, and document review easier to handle than a sequence of terminal questions.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/giannimassi/webview-cli/skill
Any agent
npx skills add giannimassi/webview-cli --skill skill
Clone the repo
git clone --depth 1 https://github.com/giannimassi/webview-cli

Made for: Claude Code, Codex.

Per session 135 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 4,368 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00135 $0.04368
Opus 5 $0.00068 $0.02184
Sonnet 5 $0.00027 $0.00874
Haiku 4.5 $0.00014 $0.00437

Measured yesterday against content hash d6d0b7107db3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

webview scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skill/SKILL.md · 268 lines

How it starts

The opening of the file, as written. The whole thing — 268 lines — stays where its author put it; the contents beside it link to each section on GitHub.

webview — Native macOS Agent UI

Open a native webview window from an agent workflow, let the user interact, get structured JSON back. Replaces terminal Q&A with a proper UI when it helps.

Execution contract [NON-NEGOTIABLE]

You — the agent — always run webview-cli via the Bash tool. Never ask the user to paste a command. The binary blocks until the user interacts with the window, then prints JSON to stdout. Your Bash call returns when the user clicks a button (or cancels/times out). Parse the result and proceed.

If running the command feels risky, ask for permission to run it — not to paste it. In auto mode, just run it.

When to use

Situation Use this skill?
Agent needs approval with context (diff, summary, risk) before a destructive action Yes — richer than [y/N]
Collecting 3+ related fields (name + role + options) Yes — one form beats N questions
Presenting 5-20 options to choose from Yes — radio/select picker
Showing a diff/preview/report before user confirms Yes — formatted content is readable
Reviewing a markdown doc (spec, PR description, draft) with comments/edits Yes — use --markdown mode (Mode 3)
Letting the user browse/edit files in a directory (tree, syntax highlighting, md preview) Yes — use --editor <path> mode (Mode 5)
Simple yes/no question No — just ask in terminal
Agent-side computation taking >30s before UI can render No — show a text status instead
Non-interactive environment (CI, piped output) No — preflight will fail, fall back to terminal

Preflight (run EVERY invocation)

# Binary on PATH?
if ! command -v webview-cli >/dev/null 2>&1; then
  echo "ERROR: webview-cli not installed. Run: brew tap giannimassi/tap && brew install webview-cli" >&2
  exit 1
fi
# macOS-only?
if [ "$(uname -s)" != "Darwin" ]; then
  echo "ERROR: webview-cli requires macOS (detected $(uname -s))." >&2
  exit 1
fi

Read the full file on GitHub · 268 lines

Files

What ships with it

3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 268 lines · 135 tokens per session scan A d6d0b7107db3

Subscribe to this mod's changes

webview is a skill published in the GitHub repository giannimassi/webview-cli (11 stars, last pushed 2mo ago), licensed MIT. It adds 135 tokens to every session and 4,368 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.