@fringetools/cli 0.1.1 → 0.1.3

package/README.md

@@ -27,7 +27,7 @@ Prefer a permanent command? `npm install -D @fringetools/cli`, then `npx fringe
 | `fringe check --hook` | One-line `fringe/hook.v1` JSON for CI / pre-commit. Exits `1` on errors. Claude Code does **not** feed this back — use `--claude-hook` for agents. Wins over `--json`. |
 | `fringe check --update-baseline` | Accept today's findings into `fringe.baseline.json` (commit it). Exit `0`. |
 | `fringe check --record [--manifest <path>]` | Also append this run to `.fringe/ledger.jsonl`, optionally merging an agent flow manifest. |
-| `fringe init [path]` | Add the Claude Code hook and install the three skills. Safe to re-run; refuses to touch an invalid `settings.json`. |
+| `fringe init [path]` | Add the Claude Code hook and install the five skills. Safe to re-run; refuses to touch an invalid `settings.json`. |
 | `fringe open [path] [--port <n>]` | Serve the ledger as a plain HTML page on `127.0.0.1:4173` (loopback only). Ctrl+C to stop. |
 | `fringe --help` · `--version` | Usage / version. |
 
@@ -101,7 +101,7 @@ After each edit the hook runs; with errors it returns, for example:
 
 This was verified end to end with real agent runs: the agent receives the findings and fixes them (see [docs/architecture.md](../../docs/architecture.md)). Why `--claude-hook` and not `--hook`: Claude Code only passes a hook's output to the agent as `decision`/`reason` JSON on exit 0 (or stderr on exit 2); the CI contract `--hook` exits 1 and is ignored.
 
-**Skills** — `fringe init` installs three into `.claude/skills/` (sources in [skill/](skill/)): `fringe` (below), `fringe-review` ("review this page": runs the checker and reports ranked, plain-English findings with fixes; invoke as `/fringe-review`) and `fringe-audit` (whole-repo plan: top rules and files, a suggested fix order, offers `--update-baseline`; `/fringe-audit`). Reviews run `fringe check`, so Claude Code will ask you to approve that command once (or allow `Bash(fringe:*)` / `Bash(npx:*)`).
+**Skills** — `fringe init` installs five into `.claude/skills/` (sources in [skill/](skill/), one folder each); the Claude Code plugin ships the same folders, namespaced `/fringe:…`: `fringe` (below), `fringe-review` ("review this page": ranked, plain-English findings with fixes), `fringe-audit` (whole-repo plan: top rules and files, a fix order, offers `--update-baseline`), `fringe-debt` (lists `fringe-ignore` suppressions and the baseline) and `fringe-help` (quick reference). Invoke as `/fringe-review` etc. Reviews run `fringe check`, so Claude Code will ask you to approve that command once (or allow `Bash(fringe:*)` / `Bash(npx:*)`).
 
 **The main skill** (`.claude/skills/fringe/SKILL.md`) teaches the agent the ladder: *Flow* (cut what users don't need, with evidence in `.fringe/manifest.json`) → *Edges* (fix every hook finding, never silence a rule to get green) → *Polish* (tokens, scales, no AI look), then record the run. It is plain Markdown, so any agent platform that reads `AGENTS.md`/skill files can use it as-is.
 

package/dist/bin.js

@@ -1,7 +1,7 @@
 #!/usr/bin/env node
 import {
   runCli
-} from "./chunk-W23H5MXS.js";
+} from "./chunk-MOTRTP2I.js";
 
 // src/bin.ts
 process.exitCode = await runCli(process.argv.slice(2));

package/dist/{chunk-W23H5MXS.js → chunk-MOTRTP2I.js}

@@ -4,7 +4,7 @@ import { readFileSync as readFileSync8 } from "fs";
 import { resolve } from "path";
 
 // src/version.ts
-var VERSION = "0.1.0";
+var VERSION = "0.1.3";
 
 // src/check.ts
 import { readFileSync as readFileSync5 } from "fs";
@@ -293,7 +293,11 @@ var fr206 = {
   severity: "error",
   check: (ctx) => {
     const out = [];
-    for (const e of ctx.elems.filter((el) => ["div", "span"].includes(el.tag) && (el.attrs.onClick ?? el.attrs.onclick) !== void 0 && el.attrs.role === void 0 && (el.attrs.tabIndex ?? el.attrs.tabindex) === void 0)) {
+    const isBackdrop = (i) => {
+      const next = ctx.elems[i + 1], me = ctx.elems[i];
+      return !!next && next.parentTag === me.tag && next.loc.line >= me.loc.line && (/^(alert)?dialog$/.test(next.attrs.role ?? "") || next.attrs["aria-modal"] === "true");
+    };
+    for (const e of ctx.elems.filter((el, i) => ["div", "span"].includes(el.tag) && (el.attrs.onClick ?? el.attrs.onclick) !== void 0 && el.attrs.role === void 0 && (el.attrs.tabIndex ?? el.attrs.tabindex) === void 0 && !isBackdrop(i))) {
       out.push({ rule: "FR-206", severity: "error", message: `onClick on non-interactive <${e.tag}>`, hint: `use a <button>/<a> or add role and tabIndex`, loc: e.loc, wcag: "2.1.1" });
     }
     return out;
@@ -1186,12 +1190,11 @@ function startBoard(root, port) {
 }
 
 // src/init.ts
-import { existsSync as existsSync2, mkdirSync as mkdirSync2, readFileSync as readFileSync7, writeFileSync as writeFileSync2, copyFileSync } from "fs";
+import { existsSync as existsSync2, mkdirSync as mkdirSync2, readFileSync as readFileSync7, readdirSync as readdirSync3, writeFileSync as writeFileSync2, copyFileSync } from "fs";
 import { dirname, join as join7 } from "path";
 import { fileURLToPath } from "url";
 var HOOK = "npx -y @fringetools/cli check --claude-hook";
-var SKILLS = { fringe: "SKILL.md", "fringe-review": "review.md", "fringe-audit": "audit.md" };
-var skillSrc = (file) => fileURLToPath(new URL(`../skill/${file}`, import.meta.url));
+var SKILL_DIR = fileURLToPath(new URL("../skill/", import.meta.url));
 function initProject(root) {
   const done = [];
   const settingsPath = join7(root, ".claude", "settings.json");
@@ -1211,12 +1214,12 @@ function initProject(root) {
     writeFileSync2(settingsPath, JSON.stringify(settings, null, 2) + "\n");
     done.push(`hook added to ${settingsPath}`);
   }
-  for (const [name, file] of Object.entries(SKILLS)) {
+  for (const name of readdirSync3(SKILL_DIR)) {
     const skillPath = join7(root, ".claude", "skills", name, "SKILL.md");
     if (existsSync2(skillPath)) done.push(`skill ${name} already present`);
     else {
       mkdirSync2(dirname(skillPath), { recursive: true });
-      copyFileSync(skillSrc(file), skillPath);
+      copyFileSync(join7(SKILL_DIR, name, "SKILL.md"), skillPath);
       done.push(`skill ${name} installed`);
     }
   }

package/dist/index.js

@@ -1,6 +1,6 @@
 import {
   runCli
-} from "./chunk-W23H5MXS.js";
+} from "./chunk-MOTRTP2I.js";
 export {
   runCli
 };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@fringetools/cli",
-  "version": "0.1.1",
+  "version": "0.1.3",
   "license": "MIT",
   "type": "module",
   "bin": {

package/skill/fringe-debt/SKILL.md

@@ -0,0 +1,17 @@
+---
+name: fringe-debt
+description: Use when the user asks what Fringe findings have been silenced or deferred ("fringe debt", "what did we suppress", "what's in the baseline", /fringe-debt). One-shot report of suppressions and the baseline; changes nothing.
+---
+
+# Fringe debt
+
+Fringe has two ways to defer a finding: an inline `fringe-ignore` comment and the `fringe.baseline.json` file. This lists both so a deferral can't quietly become permanent.
+
+Command: use `fringe` if it is on the PATH, otherwise `npx -y @fringetools/cli` in its place.
+
+1. **Inline suppressions.** `git grep --no-index --exclude-standard -nE 'fringe-ignore +FR-[0-9]+'`. One row per hit: `<file>:<line>, <rule> — <why, if the comment or nearby code says>`. Skip hits that are only prose mentioning the convention.
+2. **Baseline.** If `fringe.baseline.json` exists, count entries per rule (each entry is `rule|file|message`) and list the files with the most. If it is missing, say there is no baseline.
+3. **Today's picture.** Run `fringe check --json .` and report the current counts by severity next to the baseline total, so the user sees whether the debt is shrinking.
+4. Close with the single most valuable thing to pay down first (usually the rule with the most baselined entries that has a mechanical fix) and the command to shrink the baseline afterwards: `fringe check --update-baseline`.
+
+Change nothing. Do not run `--update-baseline` unless the user asks.

package/skill/fringe-help/SKILL.md

@@ -0,0 +1,33 @@
+---
+name: fringe-help
+description: Use when the user asks what Fringe can do or how to use it ("fringe help", "what commands does fringe have", /fringe-help). Quick reference; changes nothing.
+---
+
+# Fringe help
+
+Fringe checks agent-written frontends for accessibility, state, dark-pattern and "AI look" problems, gates the hard failures, and keeps a local history.
+
+**Skills**
+
+| Skill | Use it to |
+|---|---|
+| `fringe` | The working rules: cut what users don't need, fix every hook finding, polish, record |
+| `/fringe-review` | Review a page or component: ranked findings with plain-English fixes |
+| `/fringe-audit` | Audit the whole repo: top rules and files, a fix order, offer a baseline |
+| `/fringe-debt` | List silenced findings (`fringe-ignore`, baseline) |
+| `/fringe-help` | This reference |
+
+**Commands** (`fringe`, or `npx -y @fringetools/cli`)
+
+| Command | What it does |
+|---|---|
+| `check [dir]` | Scan and print findings and the health score; exit 1 on any error |
+| `check --json` / `--hook` / `--claude-hook` | Machine output; CI gate; agent feedback form |
+| `check --update-baseline` | Accept today's findings; only new ones gate |
+| `check --record [--manifest f]` | Append the run to `.fringe/ledger.jsonl` |
+| `init` | Add the hook and install the skills |
+| `open` | View the ledger at 127.0.0.1:4173 |
+
+**Escape hatch:** `// fringe-ignore FR-203` on or above the line, for a real false positive only; every run reports `N suppressed`.
+
+Full reference: https://github.com/skc-learn/fringe/blob/main/packages/cli/README.md