Launch One Good Commit static Space

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Joshua Bloodworth
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -1,10 +1,112 @@
 ---
 title: One Good Commit
-emoji: 📈
-colorFrom: purple
-colorTo: indigo
+emoji: 🃏
+colorFrom: yellow
+colorTo: green
 sdk: static
+app_file: index.html
 pinned: false
+license: mit
+short_description: Evidence-backed model card contribution assistant
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# One Good Commit
+
+One Good Commit reads a public Hugging Face model card and prepares a bounded
+documentation contribution packet for a human to review.
+
+It answers three narrow questions:
+
+1. What relevant documentation can be located in the public model card?
+2. Which selected documentation signals were not found by the deterministic review?
+3. What questions could a prospective contributor ask without inventing facts?
+
+The tool does **not** score compliance, certify safety, evaluate model quality,
+run the model, or post a Discussion or pull request.
+
+## Why this exists
+
+Model cards help people discover, reproduce, understand, and responsibly use
+models. Maintainers also have uneven time, information, and documentation needs.
+This project is a small contribution aid: it keeps evidence attached to each
+observation and leaves every consequential decision with a person.
+
+## Method
+
+The app retrieves two public sources:
+
+- the model repository's Hub API record;
+- the selected revision of its `README.md`.
+
+It then runs deterministic checks for documentation signals related to:
+
+- license;
+- language coverage;
+- task and model type;
+- base model or provenance;
+- intended uses;
+- out-of-scope uses;
+- limitations, risks, or biases;
+- training data;
+- evaluation evidence;
+- training and reproducibility details;
+- citations or supporting papers.
+
+A result can be:
+
+- **Documented:** relevant metadata or card text was located.
+- **Review:** related text was located, but a structured or clearer source may
+  still need human review.
+- **Not found:** the review did not locate relevant information.
+
+“Not found” is deliberately limited language. It does not mean the information
+does not exist, the section is mandatory, or the maintainer has done something
+wrong.
+
+## Deployment
+
+The public Space is intentionally static so it can run without a paid server,
+GPU, account system, or secret. Its browser-side code makes read-only requests
+to the selected public model's Hub API record and `README.md`; model-card text
+is treated as untrusted and rendered as text, not HTML.
+
+The earlier Python/Gradio prototype and fixture tests remain in this repository
+as a local reference implementation. The deployed Space is `index.html`,
+`static-app.js`, and `static-styles.css`.
+
+## Run the local reference implementation
+
+```bash
+python -m venv .venv
+source .venv/bin/activate
+pip install -r requirements.txt
+python app.py
+```
+
+The analyzer and its tests use only the Python standard library:
+
+```bash
+python -m unittest discover -s tests -v
+```
+
+## Contribution boundary
+
+Before contacting a maintainer:
+
+1. Read the complete card and linked primary sources.
+2. Confirm that the suggested section applies to that model.
+3. Remove speculative placeholders.
+4. Ask questions where only maintainers can know the answer.
+5. Prefer one small, accurate improvement over a broad rewrite.
+
+No output from this tool should be posted automatically.
+
+## Privacy and access
+
+The current version accepts public, ungated model repositories only. It does not
+accept uploaded documents or store model card content in an application
+database.
+
+## License
+
+MIT

index.html

@@ -1,19 +1,59 @@
 <!doctype html>
-<html>
-	<head>
-		<meta charset="utf-8" />
-		<meta name="viewport" content="width=device-width" />
-		<title>My static Space</title>
-		<link rel="stylesheet" href="style.css" />
-	</head>
-	<body>
-		<div class="card">
-			<h1>Welcome to your static Space!</h1>
-			<p>You can modify this app directly by editing <i>index.html</i> in the Files and versions tab.</p>
-			<p>
-				Also don't forget to check the
-				<a href="https://huggingface.co/docs/hub/spaces" target="_blank">Spaces documentation</a>.
-			</p>
-		</div>
-	</body>
+<html lang="en">
+  <head>
+    <meta charset="utf-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1">
+    <meta name="description" content="An evidence-backed, human-reviewed contribution assistant for public Hugging Face model cards.">
+    <title>One Good Commit</title>
+    <link rel="stylesheet" href="static-styles.css">
+  </head>
+  <body>
+    <main class="shell">
+      <header class="hero">
+        <p class="eyebrow">A HUMAN-REVIEWED CONTRIBUTION ASSISTANT</p>
+        <h1>One Good Commit</h1>
+        <p class="lede">Read a public Hugging Face model card, trace what it documents, and prepare a careful contribution packet for a human to review.</p>
+        <div class="principles" aria-label="Project principles">
+          <span>No compliance score</span><span>No invented details</span>
+          <span>No automatic posts</span><span>Evidence stays attached</span>
+        </div>
+      </header>
+
+      <section class="input-card" aria-label="Review a public model card">
+        <form id="review-form">
+          <label for="repo-id">Public model ID</label>
+          <div class="input-row">
+            <input id="repo-id" name="repo-id" autocomplete="off" placeholder="owner/model-name" required>
+            <button type="submit">Review card</button>
+          </div>
+        </form>
+        <p class="example-label">Try an established public card</p>
+        <div class="examples" aria-label="Example model cards">
+          <button type="button" data-example="sentence-transformers/all-MiniLM-L6-v2">sentence-transformers/all-MiniLM-L6-v2</button>
+          <button type="button" data-example="google-bert/bert-base-uncased">google-bert/bert-base-uncased</button>
+          <button type="button" data-example="openai-community/gpt2">openai-community/gpt2</button>
+        </div>
+        <p class="input-note">Public repositories only. This app reads the selected README and Hub metadata; it does not run or evaluate the model.</p>
+      </section>
+
+      <section id="notice" class="notice" hidden aria-live="polite"></section>
+      <section id="results" hidden aria-live="polite">
+        <div id="summary"></div>
+        <div class="tabs" role="tablist" aria-label="Review output">
+          <button role="tab" aria-selected="true" aria-controls="evidence-panel" id="evidence-tab" data-tab="evidence-panel">Evidence</button>
+          <button role="tab" aria-selected="false" aria-controls="worksheet-panel" id="worksheet-tab" data-tab="worksheet-panel">Contribution worksheet</button>
+          <button role="tab" aria-selected="false" aria-controls="discussion-panel" id="discussion-tab" data-tab="discussion-panel">Discussion draft</button>
+        </div>
+        <section id="evidence-panel" role="tabpanel" aria-labelledby="evidence-tab" class="tab-panel"></section>
+        <section id="worksheet-panel" role="tabpanel" aria-labelledby="worksheet-tab" class="tab-panel" hidden></section>
+        <section id="discussion-panel" role="tabpanel" aria-labelledby="discussion-tab" class="tab-panel" hidden></section>
+        <button id="download-review" class="download" type="button">Download complete review</button>
+      </section>
+
+      <footer>
+        <strong>Method boundary.</strong> This tool checks whether selected documentation signals can be located. It does not establish truth, safety, quality, legal status, or compliance. Every contribution remains a human decision.
+      </footer>
+    </main>
+    <script src="static-app.js"></script>
+  </body>
 </html>

static-app.js

@@ -0,0 +1,205 @@
+(() => {
+  "use strict";
+
+  const MAX_README_BYTES = 2_000_000;
+  const REPO_ID = /^[A-Za-z0-9][A-Za-z0-9._-]*\/[A-Za-z0-9][A-Za-z0-9._-]*$/;
+  const checks = [
+    ["license", "License", ["license", "license_name", "license_link"], [], [], "Confirm the license identifier or link that governs reuse."],
+    ["language", "Language coverage", ["language", "languages"], ["\\blanguages?\\b"], [], "Document the languages represented and any known coverage limits."],
+    ["task", "Task and model type", ["pipeline_tag", "library_name", "tags"], ["\\bmodel description\\b", "\\btask\\b"], [], "Clarify the task, model family, and expected input/output shape."],
+    ["base_model", "Base model or provenance", ["base_model"], ["\\bmodel description\\b", "\\bmodel details\\b", "\\bprovenance\\b"], ["\\bbase model\\b", "\\bfine[- ]?tuned from\\b", "\\bderived from\\b"], "Name the base model or explain the model's provenance."],
+    ["intended_uses", "Intended uses", [], ["\\bintended uses?\\b", "\\buse cases?\\b", "\\buses?\\b", "\\bhow to use\\b"], ["\\bintended for\\b", "\\bcan be used\\b", "\\buse cases?\\b"], "Describe supported use cases and the users or settings considered."],
+    ["out_of_scope", "Out-of-scope uses", [], ["\\bout[- ]of[- ]scope\\b", "\\bmisuse\\b", "\\bnot intended\\b", "\\bprohibited uses?\\b"], ["\\bshould not\\b", "\\bnot intended\\b", "\\bout[- ]of[- ]scope\\b"], "Identify unsupported or inappropriate uses, where relevant."],
+    ["limitations", "Limitations, risks, or biases", [], ["\\blimitations?\\b", "\\brisks?\\b", "\\bbias(?:es)?\\b", "\\bethical considerations?\\b"], ["\\blimitations?\\b", "\\brisks?\\b", "\\bbias(?:es|ed)?\\b"], "Describe known limitations, biases, or risks and how they were observed."],
+    ["training_data", "Training data", ["datasets", "dataset"], ["\\btraining data\\b", "\\bdatasets?\\b", "\\btraining details\\b"], ["\\btrained on\\b", "\\btraining dataset\\b"], "Name or characterize the training data and include source links when possible."],
+    ["evaluation", "Evaluation evidence", ["model-index", "eval_results", "metrics"], ["\\bevaluation\\b", "\\bresults?\\b", "\\bbenchmarks?\\b", "\\bmetrics?\\b"], ["\\bbenchmark\\b", "\\baccuracy\\b", "\\bf1\\b", "\\bbleu\\b", "\\brouge\\b"], "Link evaluation datasets, metrics, values, and their source or procedure."],
+    ["reproducibility", "Training and reproducibility details", ["training_args"], ["\\btraining procedure\\b", "\\btraining details\\b", "\\bhyperparameters?\\b", "\\breproducibility\\b"], ["\\blearning rate\\b", "\\bbatch size\\b", "\\btraining steps?\\b"], "Add the training procedure, key parameters, code, or experiment links."],
+    ["citation", "Citation or supporting paper", ["citation"], ["\\bcitation\\b", "\\breferences?\\b", "\\bpaper\\b"], [], "Add a citation, paper, technical report, or supporting reference if one exists."]
+  ];
+
+  const form = document.querySelector("#review-form");
+  const input = document.querySelector("#repo-id");
+  const notice = document.querySelector("#notice");
+  const results = document.querySelector("#results");
+  const summary = document.querySelector("#summary");
+  const evidencePanel = document.querySelector("#evidence-panel");
+  const worksheetPanel = document.querySelector("#worksheet-panel");
+  const discussionPanel = document.querySelector("#discussion-panel");
+  const downloadButton = document.querySelector("#download-review");
+  let latestReview = null;
+
+  function cleanRepoId(value) {
+    let id = value.trim();
+    if (id.startsWith("https://huggingface.co/")) id = id.slice("https://huggingface.co/".length).split("/").slice(0, 2).join("/");
+    if (!REPO_ID.test(id)) throw new Error("Enter a public model ID in the form owner/model-name.");
+    return id;
+  }
+
+  function meaningful(value) {
+    if (value === null || value === undefined || value === false) return false;
+    if (typeof value === "string") return Boolean(value.trim());
+    if (Array.isArray(value)) return value.length > 0;
+    if (typeof value === "object") return Object.keys(value).length > 0;
+    return true;
+  }
+
+  function compact(value, limit = 320) {
+    const text = (typeof value === "object" ? JSON.stringify(value) : String(value)).replace(/\s+/g, " ").trim();
+    return text.length > limit ? `${text.slice(0, limit - 1)}…` : text;
+  }
+
+  function frontmatter(readme) {
+    const lines = readme.split("\n");
+    if (lines[0]?.trim() !== "---") return {};
+    const end = lines.findIndex((line, index) => index > 0 && line.trim() === "---");
+    if (end < 0) return {};
+    const metadata = {};
+    let current = null;
+    for (const line of lines.slice(1, end)) {
+      if (!line.trim() || line.trimStart().startsWith("#")) continue;
+      const list = line.match(/^\s*-\s+(.+?)\s*$/);
+      if (list && current) { if (Array.isArray(metadata[current])) metadata[current].push(list[1].replace(/^['"]|['"]$/g, "")); continue; }
+      const key = line.match(/^([A-Za-z0-9_-]+):(?:\s*(.*))?$/);
+      if (!key) continue;
+      current = key[1];
+      const value = (key[2] || "").trim();
+      metadata[current] = value ? value.replace(/^['"]|['"]$/g, "") : [];
+    }
+    return metadata;
+  }
+
+  function sections(readme) {
+    const lines = readme.split("\n");
+    const headings = [];
+    let fenced = false;
+    lines.forEach((line, index) => {
+      if (/^\s*(```|~~~)/.test(line)) { fenced = !fenced; return; }
+      if (fenced) return;
+      const match = line.match(/^(#{1,6})\s+(.+?)\s*#*\s*$/);
+      if (match) headings.push({ start: index + 1, level: match[1].length, title: match[2].trim() });
+    });
+    return headings.map((heading, index) => {
+      let end = lines.length;
+      for (const next of headings.slice(index + 1)) { if (next.level <= heading.level) { end = next.start - 1; break; } }
+      return { ...heading, body: lines.slice(heading.start, end).join("\n").trim() };
+    });
+  }
+
+  function sourceLink(repo, revision, line) {
+    const link = `https://huggingface.co/${repo}/blob/${revision}/README.md`;
+    return line ? `${link}#L${line}` : link;
+  }
+
+  function analyze(repo, api, readme) {
+    const revision = api.sha || "main";
+    const metadata = { ...frontmatter(readme), ...(api.cardData || {}) };
+    if (!metadata.pipeline_tag && api.pipeline_tag) metadata.pipeline_tag = api.pipeline_tag;
+    if (!metadata.library_name && api.library_name) metadata.library_name = api.library_name;
+    if (!metadata.tags && api.tags) metadata.tags = api.tags;
+    const parsedSections = sections(readme);
+    const lines = readme.split("\n");
+    const findings = checks.map(([key, label, keys, headings, keywords, prompt]) => {
+      const evidence = [];
+      const foundKeys = keys.filter((item) => meaningful(metadata[item]));
+      foundKeys.slice(0, 2).forEach((item) => evidence.push({ type: "Model card metadata", text: `${item}: ${compact(metadata[item])}`, url: sourceLink(repo, revision) }));
+      const matchedSections = parsedSections.filter((section) => headings.some((pattern) => new RegExp(pattern, "i").test(section.title)));
+      if (matchedSections[0]) {
+        const item = matchedSections[0];
+        evidence.push({ type: "Model card section", text: compact(`${item.title}: ${item.body}`, 420), url: sourceLink(repo, revision, item.start) });
+      }
+      if (!evidence.length && keywords.length) {
+        const foundLine = lines.findIndex((line) => line.trim() && !line.trim().startsWith("#") && keywords.some((pattern) => new RegExp(pattern, "i").test(line)));
+        if (foundLine >= 0) evidence.push({ type: "Model card text", text: compact(lines[foundLine], 420), url: sourceLink(repo, revision, foundLine + 1) });
+      }
+      let status = evidence.length ? "documented" : "not_found";
+      let text = evidence.length ? "Relevant information was found in the public card." : "Relevant information was not found by this review.";
+      if (key === "evaluation" && matchedSections.length && !foundKeys.length) { status = "review"; text = "Evaluation text was found, but structured evaluation metadata was not found."; }
+      if (key === "task" && foundKeys.length === 1 && foundKeys[0] === "tags") { status = "review"; text = "Relevant tags were found; the narrative description may still need review."; }
+      return { key, label, status, text, evidence, prompt };
+    });
+    return { repo, revision, sourceUrl: sourceLink(repo, revision), findings, analyzedAt: new Date().toISOString() };
+  }
+
+  function counts(review) { return review.findings.reduce((all, finding) => ({ ...all, [finding.status]: (all[finding.status] || 0) + 1 }), { documented: 0, review: 0, not_found: 0 }); }
+  function statusLabel(status) { return status === "not_found" ? "Not found" : status[0].toUpperCase() + status.slice(1); }
+  function make(tag, className, text) { const element = document.createElement(tag); if (className) element.className = className; if (text !== undefined) element.textContent = text; return element; }
+
+  function renderSummary(review) {
+    const count = counts(review);
+    summary.replaceChildren();
+    const header = make("section", "result-header");
+    const heading = make("div");
+    heading.append(make("p", "eyebrow", "PUBLIC MODEL CARD REVIEW"), make("h2", "", review.repo));
+    const link = make("a", "", "Open reviewed revision ↗"); link.href = review.sourceUrl; link.target = "_blank"; link.rel = "noopener"; heading.append(link);
+    const strip = make("div", "count-strip");
+    [[count.documented, "documented"], [count.review, "review"], [count.not_found, "not found"]].forEach(([number, label]) => { const entry = make("div"); entry.append(make("strong", "", number), make("span", "", label)); strip.append(entry); });
+    header.append(heading, strip);
+    const note = make("p", "scope-note", "A narrow result: “Not found” means the review did not locate relevant information. It is not a compliance grade or a claim of maintainer error.");
+    const grid = make("section", "finding-grid");
+    review.findings.forEach((finding) => { const card = make("article", `finding-card ${finding.status.replace("_", "-")}`); const top = make("div", "finding-topline"); top.append(make("h3", "", finding.label), make("span", "status-pill", statusLabel(finding.status))); card.append(top, make("p", "", finding.text), make("small", "", finding.evidence.length ? `${finding.evidence.length} source item${finding.evidence.length === 1 ? "" : "s"}` : "Maintainer input needed")); grid.append(card); });
+    summary.append(header, note, grid);
+  }
+
+  function renderEvidence(review) {
+    evidencePanel.replaceChildren();
+    review.findings.forEach((finding) => { const item = make("article", "evidence-item"); item.append(make("h3", "", `${finding.label} — ${statusLabel(finding.status)}`), make("p", "", finding.text)); if (finding.evidence.length) { finding.evidence.forEach((evidence) => { const link = make("a", "", evidence.type); link.href = evidence.url; link.target = "_blank"; link.rel = "noopener"; const excerpt = make("div", "excerpt", evidence.text); item.append(link, excerpt); }); } else { item.append(make("p", "", `Question for the maintainer: ${finding.prompt}`)); } evidencePanel.append(item); });
+  }
+
+  function worksheet(review) {
+    const unresolved = review.findings.filter((finding) => finding.status !== "documented");
+    const lines = [`# Suggested model card additions for ${review.repo}`, "", "> Review worksheet—not a drop-in patch. Replace every TODO with maintainer-confirmed information.", ""];
+    if (!unresolved.length) return lines.concat(["The selected checks found relevant documentation for every reviewed area.", "A human should still verify accuracy, currency, and applicability."]).join("\n");
+    unresolved.forEach((finding) => lines.push(`## ${finding.label}`, "", `<!-- TODO: ${finding.prompt} -->`, ""));
+    return lines.concat(["## Review notes", "", "- Confirm that every proposed addition is accurate with the maintainer.", "- Link primary datasets, evaluations, papers, and code where available.", "- Remove sections that are not applicable instead of filling them speculatively."]).join("\n");
+  }
+
+  function discussion(review) {
+    const unresolved = review.findings.filter((finding) => finding.status !== "documented");
+    const labels = unresolved.slice(0, 4).map((finding) => finding.label.toLowerCase()).join(", ");
+    const observation = unresolved.length ? `While reading the public card, I could not confidently locate documentation for ${labels}${unresolved.length > 4 ? `, and ${unresolved.length - 4} other area(s)` : ""}. These may be in another source or may not apply to this model.` : "The review found relevant documentation for each checked area. I would still appreciate confirmation that the information is current.";
+    return ["## Possible documentation contribution", "", `Hi—thank you for sharing \`${review.repo}\`.`, "", observation, "", "Would a small documentation contribution be useful? I would be glad to prepare one, but I do not want to infer training, evaluation, licensing, or intended-use details that only the maintainers can confirm.", "", `Card reviewed: ${review.sourceUrl}`, "", "_Drafted for human review. This message has not been posted._"].join("\n");
+  }
+
+  function fullReview(review, worksheetText, discussionText) {
+    const count = counts(review);
+    const lines = [`# One Good Commit review: ${review.repo}`, "", `- Source: ${review.sourceUrl}`, `- Revision: \`${review.revision}\``, `- Analyzed: ${review.analyzedAt}`, `- Documented: ${count.documented}`, `- Review: ${count.review}`, `- Not found: ${count.not_found}`, "", "> “Not found” means this deterministic review did not locate relevant information. It is not a claim of noncompliance, inaccuracy, or maintainer error.", "", "## Evidence", ""];
+    review.findings.forEach((finding) => { lines.push(`### ${finding.label} — ${statusLabel(finding.status)}`, "", finding.text, ""); if (finding.evidence.length) finding.evidence.forEach((evidence) => lines.push(`- **${evidence.type}** ([source](${evidence.url}))`, "", `  > ${evidence.text}`, "")); else lines.push(`- Maintainer question: ${finding.prompt}`, ""); });
+    return lines.concat(["---", "", worksheetText, "", "---", "", discussionText]).join("\n");
+  }
+
+  function showError(message) { notice.textContent = message; notice.hidden = false; results.hidden = true; }
+  function clearNotice() { notice.hidden = true; notice.textContent = ""; }
+
+  async function review(value) {
+    let repo;
+    try { repo = cleanRepoId(value); } catch (error) { showError(error.message); return; }
+    clearNotice(); results.hidden = true;
+    const button = form.querySelector("button"); button.disabled = true; button.textContent = "Reading public card…";
+    try {
+      const encoded = repo.split("/").map(encodeURIComponent).join("/");
+      const apiResponse = await fetch(`https://huggingface.co/api/models/${encoded}`);
+      if (apiResponse.status === 404) throw new Error("That public model repository was not found.");
+      if (!apiResponse.ok) throw new Error("The Hub API could not be reached. Try again shortly.");
+      const api = await apiResponse.json();
+      if (api.private || api.gated) throw new Error("That model is private or gated. This version reviews public cards only.");
+      const revision = api.sha || "main";
+      const cardResponse = await fetch(`https://huggingface.co/${encoded}/resolve/${encodeURIComponent(revision)}/README.md`);
+      if (!cardResponse.ok) throw new Error("The repository does not expose a README.md on its selected revision.");
+      const readme = await cardResponse.text();
+      if (readme.length > MAX_README_BYTES) throw new Error("The model card is larger than the 2 MB review limit.");
+      latestReview = analyze(repo, api, readme);
+      const worksheetText = worksheet(latestReview); const discussionText = discussion(latestReview);
+      renderSummary(latestReview); renderEvidence(latestReview);
+      worksheetPanel.replaceChildren(make("pre", "worksheet", worksheetText));
+      discussionPanel.replaceChildren(make("pre", "discussion", discussionText));
+      latestReview.download = fullReview(latestReview, worksheetText, discussionText);
+      results.hidden = false;
+    } catch (error) { showError(error.message || "An unexpected error occurred. No contribution was created or posted."); }
+    finally { button.disabled = false; button.textContent = "Review card"; }
+  }
+
+  form.addEventListener("submit", (event) => { event.preventDefault(); review(input.value); });
+  document.querySelectorAll("[data-example]").forEach((button) => button.addEventListener("click", () => { input.value = button.dataset.example; input.focus(); }));
+  document.querySelectorAll("[role=tab]").forEach((tab) => tab.addEventListener("click", () => { document.querySelectorAll("[role=tab]").forEach((item) => item.setAttribute("aria-selected", String(item === tab))); document.querySelectorAll("[role=tabpanel]").forEach((panel) => { panel.hidden = panel.id !== tab.dataset.tab; }); }));
+  downloadButton.addEventListener("click", () => { if (!latestReview?.download) return; const blob = new Blob([latestReview.download], { type: "text/markdown;charset=utf-8" }); const link = document.createElement("a"); link.href = URL.createObjectURL(blob); link.download = `${latestReview.repo.replace("/", "--")}-review.md`; link.click(); URL.revokeObjectURL(link.href); });
+})();

static-styles.css

@@ -0,0 +1 @@
+:root{--ink:#172419;--muted:#647067;--paper:#fbfaf4;--line:#dfe4da;--green:#2d6a4f;--green-soft:#e7f2e9;--amber:#9a6700;--amber-soft:#fff3cf;--rose:#9f3a38;--rose-soft:#fbe8e4}*{box-sizing:border-box}body{margin:0;background:radial-gradient(circle at 12% 8%,rgba(255,214,102,.18),transparent 24rem),var(--paper);color:var(--ink);font-family:Inter,ui-sans-serif,system-ui,-apple-system,BlinkMacSystemFont,"Segoe UI",sans-serif}.shell{max-width:1080px;margin:0 auto;padding:30px 18px 56px}.hero{padding:30px 0 20px;max-width:820px}.eyebrow{color:var(--green);font-size:.75rem;font-weight:800;letter-spacing:.12em;margin:0 0 10px}.hero h1{font-family:Georgia,serif;font-size:clamp(2.5rem,7vw,5.5rem);line-height:.98;letter-spacing:-.05em;margin:0}.lede{color:#435047;font-size:1.1rem;line-height:1.65;max-width:720px;margin:20px 0 10px}.principles{display:flex;flex-wrap:wrap;gap:8px;margin-top:20px}.principles span{border:1px solid var(--line);border-radius:999px;padding:7px 11px;background:rgba(255,255,255,.7);font-size:.82rem}.input-card{border:1px solid var(--line);border-radius:18px;background:rgba(255,255,255,.82);padding:18px;box-shadow:0 16px 50px rgba(35,50,39,.07)}label{display:block;font-weight:700;margin-bottom:8px}.input-row{display:flex;gap:12px}input{min-width:0;flex:1;border:1px solid #d6d9d4;border-radius:8px;padding:13px 14px;font:inherit;color:var(--ink)}button{font:inherit;cursor:pointer}form button{background:var(--ink);border:1px solid var(--ink);border-radius:8px;color:#fff;font-weight:750;padding:0 21px}.example-label{color:var(--muted);font-size:.83rem;margin:15px 0 8px}.examples{display:flex;flex-wrap:wrap;gap:8px}.examples button{border:1px solid var(--line);border-radius:7px;background:#fff;padding:7px 10px;color:var(--ink);font-size:.82rem}.input-note{color:var(--muted);font-size:.83rem;margin:14px 0 0}.notice{border:1px solid #eab4ac;background:var(--rose-soft);color:#702e2c;padding:16px;border-radius:12px;margin-top:18px}.result-header{display:flex;gap:24px;align-items:end;justify-content:space-between;padding:28px 0 16px;border-bottom:1px solid var(--line)}.result-header h2{font-family:Georgia,serif;font-size:clamp(1.6rem,4vw,2.5rem);margin:0 0 6px;overflow-wrap:anywhere}.result-header a{color:var(--green);font-weight:650;text-decoration:none}.count-strip{display:flex;gap:10px}.count-strip div{min-width:88px;border-left:1px solid var(--line);padding-left:12px}.count-strip strong{display:block;font-family:Georgia,serif;font-size:1.8rem}.count-strip span{display:block;color:var(--muted);font-size:.72rem;text-transform:uppercase;letter-spacing:.06em}.scope-note{background:#f1f0e8;border-radius:12px;padding:13px 15px;color:#4f5b52;font-size:.88rem;margin:18px 0}.finding-grid{display:grid;grid-template-columns:repeat(2,minmax(0,1fr));gap:12px}.finding-card{background:#fff;border:1px solid var(--line);border-left:4px solid var(--line);border-radius:12px;padding:15px}.finding-card.documented{border-left-color:var(--green)}.finding-card.review{border-left-color:var(--amber)}.finding-card.not-found{border-left-color:var(--rose)}.finding-topline{display:flex;gap:12px;align-items:start;justify-content:space-between}.finding-card h3{font-size:1rem;margin:0}.finding-card p{color:#536057;line-height:1.5;min-height:2.9em}.finding-card small{color:var(--muted)}.status-pill{border-radius:999px;padding:4px 8px;font-size:.7rem;font-weight:800;white-space:nowrap}.documented .status-pill{color:var(--green);background:var(--green-soft)}.review .status-pill{color:var(--amber);background:var(--amber-soft)}.not-found .status-pill{color:var(--rose);background:var(--rose-soft)}.tabs{display:flex;gap:20px;border-bottom:1px solid var(--line);margin-top:26px}.tabs button{background:none;border:0;border-bottom:2px solid transparent;padding:10px 0;color:#445147}.tabs button[aria-selected="true"]{border-color:#f97316;color:#c2410c}.tab-panel{padding:18px 0}.evidence-item{border-bottom:1px solid var(--line);padding:14px 0}.evidence-item h3{margin:0 0 6px}.evidence-item p{line-height:1.55}.evidence-item a{color:var(--green);font-weight:650}.excerpt{background:#f4f5f0;border-left:3px solid #d4d9d0;border-radius:0 6px 6px 0;color:#3f4a42;font-family:ui-monospace,SFMono-Regular,Menlo,monospace;font-size:.83rem;line-height:1.55;margin:10px 0 0;padding:10px;white-space:pre-wrap;overflow-wrap:anywhere}.worksheet,.discussion{background:#f4f5f0;border:1px solid var(--line);border-radius:10px;line-height:1.6;margin:0;overflow:auto;padding:16px;white-space:pre-wrap}.download{border:1px solid var(--line);border-radius:8px;background:#fff;color:var(--ink);font-weight:700;padding:10px 13px}footer{color:var(--muted);font-size:.82rem;line-height:1.6;padding:26px 0 0}@media(max-width:720px){.result-header{display:block}.count-strip{margin-top:20px}.count-strip div{min-width:0;flex:1}.finding-grid{grid-template-columns:1fr}.input-row{display:block}.input-row button{height:46px;margin-top:10px;width:100%}.tabs{gap:12px;overflow:auto}.tabs button{font-size:.83rem;white-space:nowrap}}