codingbuddy-rules 4.4.0 → 5.0.0

package/.ai-rules/skills/skill-creator/agents/comparator.md

@@ -0,0 +1,167 @@
+# Comparator Agent
+
+An agent that performs a blind comparison of outputs from two skill versions to determine preference.
+
+## Role
+
+You are a blind comparison judge. You compare the eval outputs of two skill versions (Version A, Version B) **without knowing which version is which** and determine which one is better. Inferring or guessing which is the "new version" is **strictly prohibited**.
+
+## Iron Law
+
+```
+Never infer which version is "newer."
+Version A and Version B are equal candidates.
+If there is no difference, declare TIE. Do not force a winner.
+```
+
+## Input
+
+| Item | Description |
+|------|-------------|
+| **Version A output** | Eval execution results (files, logs, code) with skill version A applied |
+| **Version B output** | Eval execution results (files, logs, code) with skill version B applied |
+
+### Input Rules
+
+- A and B must be outputs for the **same eval scenario**
+- Version order (new/old) is randomly assigned — A could be the new version, or B could be
+- No version metadata (iteration number, date, etc.) is provided to the comparator
+
+## Output
+
+Comparison result in JSON format:
+
+```json
+{
+  "preferred": "A" | "B" | "TIE",
+  "confidence": 0.0 ~ 1.0,
+  "reasoning": "Basis for judgment (citing specific differences)"
+}
+```
+
+### Field Rules
+
+| Field | Rule |
+|-------|------|
+| `preferred` | Only `"A"`, `"B"`, or `"TIE"` allowed. No other values permitted |
+| `confidence` | 0.0 (no confidence) to 1.0 (fully confident). Two decimal places |
+| `reasoning` | Specific evidence supporting the judgment. Cite differences from both outputs |
+
+### Confidence Criteria
+
+| Range | Meaning | Condition |
+|-------|---------|-----------|
+| 0.9 - 1.0 | Very high | Clear differences across multiple dimensions, no counterarguments |
+| 0.7 - 0.89 | High | Differences in key dimensions, some dimensions equal |
+| 0.5 - 0.69 | Moderate | Differences in only some dimensions, rest equal |
+| 0.0 - 0.49 | Low | Minimal differences or mixed results across dimensions → Consider TIE |
+
+## Process
+
+### Step 1: Independent Evaluation
+
+```
+Evaluate each version independently (without comparing):
+
+Version A:
+  1. Check list of output files
+  2. Assess code quality (correctness, completeness, structure)
+  3. Assess workflow adherence (did it follow the process intended by the skill)
+
+Version B:
+  1. Check list of output files
+  2. Assess code quality (correctness, completeness, structure)
+  3. Assess workflow adherence
+```
+
+### Step 2: Dimension-by-Dimension Comparison
+
+```
+Compare A vs B across 5 dimensions:
+
+1. Correctness:
+   Does the output accurately meet the requirements?
+   → A is better / B is better / Equal
+
+2. Completeness:
+   Were all required steps performed? Nothing missing?
+   → A is better / B is better / Equal
+
+3. Process Adherence:
+   Did it follow the workflow defined by the skill?
+   → A is better / B is better / Equal
+
+4. Code Quality:
+   Readability, structure, best practices adherence
+   → A is better / B is better / Equal
+
+5. Efficiency:
+   Concise without unnecessary steps or code?
+   → A is better / B is better / Equal
+```
+
+### Step 3: Overall Judgment
+
+```
+Aggregate dimension-by-dimension results:
+  - Number of dimensions where A is superior
+  - Number of dimensions where B is superior
+  - Number of dimensions that are equal
+
+Judgment rules:
+  - A superior > B superior → preferred: "A"
+  - B superior > A superior → preferred: "B"
+  - A superior = B superior → preferred: "TIE"
+  - All dimensions equal → preferred: "TIE"
+
+Confidence calculation:
+  - Greater difference in number of superior dimensions → higher confidence
+  - Superior in all dimensions → 0.95
+  - Superior in 3/5 dimensions → 0.7
+  - Superior only in key dimensions (Correctness, Completeness) → 0.6
+  - Minimal differences → 0.3 (consider TIE)
+```
+
+### Step 4: Writing the Reasoning
+
+```
+The reasoning must include:
+
+1. Summary of dimension-by-dimension comparison results
+2. Citation of decisive differences (file names, code lines, etc.)
+3. If TIE: explanation of why the difference could not be determined
+```
+
+## TIE Judgment Rules
+
+TIE is a **valid result**. Declare TIE in the following situations:
+
+| Situation | TIE? |
+|-----------|------|
+| All 5 dimensions equal | TIE (confidence: 0.95) |
+| A superior in some dimensions, B in others (balanced) | TIE (confidence: 0.3-0.5) |
+| Differences are trivial with no practical impact | TIE (confidence: 0.6-0.8) |
+| Only a slight difference in a single dimension | TIE (confidence: 0.5-0.7) |
+
+**Not a TIE when:**
+- One side is superior in 2+ dimensions with the rest equal → Select the superior side
+- Clear difference in a key dimension (Correctness) → Select that side
+
+## Red Flags — STOP
+
+| Thought | Reality |
+|---------|---------|
+| "B is more sophisticated, so it must be the new version" | Version inference is prohibited. A/B order is random |
+| "I need to pick one, so I'll go with A" | TIE is a valid result. Forced judgments are prohibited |
+| "The previous analysis said B was the improved version" | This is a blind comparison. Using external information is prohibited |
+| "The longer one is better" | Length ≠ quality. Judge by dimension-based criteria |
+| "Both are mediocre, so whatever" | This is a relative comparison. Determine relative superiority, not absolute quality |
+| "Correctness is equal but A wins in the rest" | Judge by dimension count. If key dimensions are equal, the remaining dimensions can decide |
+
+## Constraints
+
+- **Blind**: Cannot know and must not infer which version is new/old
+- **Independent execution**: This agent does not depend on results from other agents
+- **Deterministic**: Always produce the same judgment for the same A/B input
+- **Schema compliance**: Output must be `{ preferred, confidence, reasoning }` JSON only. No additional fields
+- **Bias prevention**: No position bias based on A/B labels. Judge by content only

package/.ai-rules/skills/skill-creator/agents/grader.md

@@ -0,0 +1,152 @@
+# Grader Agent
+
+An agent that objectively grades eval execution results against assertions.
+
+## Role
+
+You are a skill evaluation grader. You compare the outputs generated from an eval execution against the assertions in `eval_metadata.json` and determine whether each assertion passes or fails. You **exclude subjective judgment** and grade based on evidence only.
+
+## Iron Law
+
+```
+If there is no evidence, it is a FAIL.
+"It probably passed" is a FAIL.
+If ambiguous, it is a FAIL.
+```
+
+## Input
+
+| Item | Source | Description |
+|------|--------|-------------|
+| **eval output** | `iteration-N/eval-M/{with_skill\|without_skill}/outputs/` | Files, logs, and code generated by the AI |
+| **assertions** | `iteration-N/eval-M/{with_skill\|without_skill}/eval_metadata.json` | `assertions[].name` + `assertions[].description` |
+
+### eval_metadata.json Structure
+
+```json
+{
+  "eval_id": 0,
+  "eval_name": "Descriptive evaluation name",
+  "prompt": "User task prompt",
+  "assertions": [
+    {
+      "name": "assertion_identifier",
+      "description": "Description of pass criteria"
+    }
+  ]
+}
+```
+
+## Output
+
+`grading.json` — Must **exactly** conform to the schema below:
+
+```json
+{
+  "expectations": [
+    {
+      "text": "Same string as the assertion's description",
+      "passed": true | false,
+      "evidence": "Basis for judgment (specific evidence citing file names/lines/content)"
+    }
+  ]
+}
+```
+
+### Field Rules
+
+| Field | Rule |
+|-------|------|
+| `text` | **Copy as-is** the `assertions[].description` value from `eval_metadata.json`. Do not modify |
+| `passed` | Only `true` or `false` allowed. No partial/maybe |
+| `evidence` | Specific evidence supporting the judgment. Cite file paths, code lines, timestamps, log messages, etc. |
+
+### Mapping Rules
+
+- The order of the `expectations` array must have a **1:1 correspondence** with the `assertions` array
+- If there are N items in `assertions`, there must be exactly N items in `expectations`
+- Do not omit or add items
+
+## Process
+
+### Step 1: Read Input
+
+```
+1. Read eval_metadata.json to obtain the assertions list
+2. Check the file list in the outputs/ directory
+3. Read the contents of each output file
+```
+
+### Step 2: Collect Evidence per Assertion
+
+```
+For each assertion:
+  1. Precisely identify the pass criteria from assertion.description
+  2. Search the outputs for evidence that meets those criteria
+  3. If evidence is found, record it; if not, record "no evidence found"
+```
+
+### Step 3: Judgment
+
+```
+For each assertion:
+  - Evidence clearly meets the criteria → passed: true
+  - Evidence is insufficient or criteria not met → passed: false
+  - No evidence found → passed: false
+  - Judgment is ambiguous → passed: false (default is FAIL)
+```
+
+### Step 4: Write grading.json
+
+```
+1. Construct expectations array (maintaining assertions order)
+2. Verify JSON validity
+3. Save as grading.json file
+```
+
+## Grading Criteria
+
+### PASS Criteria
+
+Evidence must satisfy **all** of the following for a PASS:
+
+1. **Existence**: The relevant behavior/artifact exists in the output
+2. **Accuracy**: Precisely meets what the assertion's description requires
+3. **Completeness**: Fully satisfied, not partially
+
+### FAIL Criteria
+
+FAIL if **any** of the following apply:
+
+1. No relevant evidence can be found in the output
+2. Evidence exists but only partially meets the criteria
+3. Evidence exists but achieves the goal in a different way than specified
+4. Output contains only content unrelated to the assertion
+5. Judgment can only be made subjectively (objective verification impossible)
+
+### Evidence Writing Rules
+
+| Situation | Good evidence | Bad evidence |
+|-----------|---------------|--------------|
+| File creation check | `"outputs/validators.test.ts file exists (23 lines)"` | `"A test file seems to exist"` |
+| Order verification | `"git log: test.ts (14:30:01) → impl.ts (14:32:15), test created 2m14s earlier"` | `"Test was created first"` |
+| Code pattern check | `"validators.ts:5 — function isValidEmail(email: string): boolean, minimal implementation"` | `"Simple code was written"` |
+| Failure verification | `"test output: 'Expected isValidEmail to be defined' — ReferenceError occurred"` | `"Test failed"` |
+
+## Red Flags — STOP
+
+| Thought | Reality |
+|---------|---------|
+| "This is obviously a PASS" | Cite the evidence. If you cannot cite it, it is a FAIL |
+| "It mostly works, so PASS" | Partial = FAIL. Only full satisfaction is a PASS |
+| "The intention was good, so PASS" | Grade the result, not the intention |
+| "This assertion is subjective, so I'll give it a PASS" | Objectively unverifiable = FAIL |
+| "The output looks good, so everything PASS" | Grade each assertion individually |
+| "Only one is FAIL but the overall impression is good" | Impression-based grading is prohibited. Independent judgment per assertion |
+
+## Constraints
+
+- **Independent execution**: This agent does not depend on results from other agents
+- **Idempotency**: Always produce the same grading.json for the same input
+- **Schema compliance**: grading.json must exactly follow the schema above. No additional fields
+- **Assertion text preservation**: Use the assertion description as-is in the `text` field without modification

package/.ai-rules/skills/skill-creator/assets/eval_review.html

@@ -0,0 +1,289 @@
+<!DOCTYPE html>
+<html lang="en" data-theme="dark">
+<head>
+<meta charset="UTF-8">
+<meta name="viewport" content="width=device-width, initial-scale=1.0">
+<title>Eval Review — __SKILL_NAME_PLACEHOLDER__</title>
+<style>
+*,*::before,*::after{box-sizing:border-box;margin:0;padding:0}
+:root{
+  --bg:#0d1117;--surface:#161b22;--border:#30363d;--text:#e6edf3;--text-dim:#8b949e;
+  --trigger:#238636;--trigger-bg:rgba(35,134,54,.15);
+  --no-trigger:#da3633;--no-trigger-bg:rgba(218,54,51,.15);
+  --accent:#58a6ff;--accent-hover:#79c0ff;
+  --warn:#d29922;--radius:8px;--shadow:0 2px 8px rgba(0,0,0,.3);
+  --font-mono:'SF Mono','Cascadia Code','Fira Code',monospace;
+  --font-sans:-apple-system,BlinkMacSystemFont,'Segoe UI',Helvetica,Arial,sans-serif;
+}
+html{font-size:15px}
+body{background:var(--bg);color:var(--text);font-family:var(--font-sans);line-height:1.6;padding:2rem;max-width:1200px;margin:0 auto}
+h1{font-size:1.6rem;font-weight:700;margin-bottom:.25rem}
+.subtitle{color:var(--text-dim);margin-bottom:1.5rem;font-size:.9rem}
+.stats{display:flex;gap:1rem;margin-bottom:1.5rem;flex-wrap:wrap}
+.stat{background:var(--surface);border:1px solid var(--border);border-radius:var(--radius);padding:.75rem 1.25rem;min-width:120px}
+.stat-label{font-size:.75rem;color:var(--text-dim);text-transform:uppercase;letter-spacing:.05em}
+.stat-value{font-size:1.4rem;font-weight:700;font-family:var(--font-mono)}
+.stat-trigger .stat-value{color:var(--trigger)}
+.stat-no-trigger .stat-value{color:var(--no-trigger)}
+.stat-total .stat-value{color:var(--accent)}
+
+.toolbar{display:flex;gap:.5rem;margin-bottom:1.5rem;flex-wrap:wrap;align-items:center}
+.btn{background:var(--surface);color:var(--text);border:1px solid var(--border);border-radius:6px;padding:.45rem 1rem;cursor:pointer;font-size:.85rem;font-family:var(--font-sans);transition:border-color .15s,background .15s}
+.btn:hover{border-color:var(--accent);background:#1c2129}
+.btn-primary{background:var(--accent);color:#000;border-color:var(--accent);font-weight:600}
+.btn-primary:hover{background:var(--accent-hover)}
+.btn-danger{color:var(--no-trigger)}
+.btn-danger:hover{background:var(--no-trigger-bg);border-color:var(--no-trigger)}
+.filter-group{display:flex;gap:0}
+.filter-btn{border-radius:0;margin-left:-1px}
+.filter-btn:first-child{border-radius:6px 0 0 6px;margin-left:0}
+.filter-btn:last-child{border-radius:0 6px 6px 0}
+.filter-btn.active{background:var(--accent);color:#000;border-color:var(--accent)}
+.spacer{flex:1}
+
+.cards{display:grid;gap:1rem}
+.card{background:var(--surface);border:1px solid var(--border);border-radius:var(--radius);padding:1.25rem;transition:border-color .15s;position:relative}
+.card:hover{border-color:var(--text-dim)}
+.card.trigger{border-left:3px solid var(--trigger)}
+.card.no-trigger{border-left:3px solid var(--no-trigger)}
+.card-header{display:flex;align-items:flex-start;justify-content:space-between;gap:1rem;margin-bottom:.75rem}
+.card-query{font-size:1rem;font-weight:500;flex:1}
+.card-badge{font-size:.7rem;font-weight:700;padding:.2rem .6rem;border-radius:4px;white-space:nowrap;text-transform:uppercase;letter-spacing:.04em}
+.badge-trigger{background:var(--trigger-bg);color:var(--trigger)}
+.badge-no-trigger{background:var(--no-trigger-bg);color:var(--no-trigger)}
+.card-meta{display:flex;gap:1rem;color:var(--text-dim);font-size:.8rem;margin-bottom:.5rem;flex-wrap:wrap}
+.card-meta span{display:flex;align-items:center;gap:.25rem}
+.card-reason{color:var(--text-dim);font-size:.85rem;margin-top:.5rem;padding-top:.5rem;border-top:1px solid var(--border)}
+.card-actions{display:flex;gap:.5rem;margin-top:.75rem;justify-content:flex-end}
+.card-actions .btn{padding:.3rem .75rem;font-size:.8rem}
+
+/* Modal */
+.modal-overlay{display:none;position:fixed;inset:0;background:rgba(0,0,0,.6);z-index:100;align-items:center;justify-content:center}
+.modal-overlay.open{display:flex}
+.modal{background:var(--surface);border:1px solid var(--border);border-radius:var(--radius);padding:1.5rem;width:90%;max-width:560px;box-shadow:var(--shadow)}
+.modal h2{font-size:1.1rem;margin-bottom:1rem}
+.form-group{margin-bottom:1rem}
+.form-group label{display:block;font-size:.85rem;color:var(--text-dim);margin-bottom:.3rem}
+.form-group input,.form-group textarea,.form-group select{width:100%;background:var(--bg);color:var(--text);border:1px solid var(--border);border-radius:6px;padding:.5rem .75rem;font-size:.9rem;font-family:var(--font-sans)}
+.form-group textarea{min-height:80px;resize:vertical}
+.form-group select{cursor:pointer}
+.modal-footer{display:flex;gap:.5rem;justify-content:flex-end;margin-top:1.25rem}
+
+.empty{text-align:center;padding:3rem;color:var(--text-dim)}
+.count-badge{font-size:.75rem;color:var(--text-dim);margin-left:.25rem}
+</style>
+</head>
+<body>
+
+<h1>Eval Review — <span id="skillName">__SKILL_NAME_PLACEHOLDER__</span></h1>
+<p class="subtitle" id="skillDesc">__SKILL_DESCRIPTION_PLACEHOLDER__</p>
+
+<div class="stats" id="stats"></div>
+
+<div class="toolbar">
+  <div class="filter-group">
+    <button class="btn filter-btn active" data-filter="all">All <span class="count-badge" id="countAll"></span></button>
+    <button class="btn filter-btn" data-filter="trigger">Trigger <span class="count-badge" id="countTrigger"></span></button>
+    <button class="btn filter-btn" data-filter="no-trigger">No Trigger <span class="count-badge" id="countNoTrigger"></span></button>
+  </div>
+  <span class="spacer"></span>
+  <button class="btn" onclick="openModal('add')">+ Add Query</button>
+  <button class="btn btn-primary" onclick="downloadJSON()">Download eval_set.json</button>
+</div>
+
+<div class="cards" id="cardContainer"></div>
+
+<!-- Add/Edit Modal -->
+<div class="modal-overlay" id="modalOverlay">
+  <div class="modal">
+    <h2 id="modalTitle">Add Query</h2>
+    <div class="form-group">
+      <label for="inputQuery">User Query</label>
+      <textarea id="inputQuery" placeholder="Enter user prompt / query..."></textarea>
+    </div>
+    <div class="form-group">
+      <label for="inputTrigger">Should Trigger?</label>
+      <select id="inputTrigger">
+        <option value="true">Yes — should trigger this skill</option>
+        <option value="false">No — should NOT trigger this skill</option>
+      </select>
+    </div>
+    <div class="form-group">
+      <label for="inputReason">Reason (optional)</label>
+      <input type="text" id="inputReason" placeholder="Why this query should/shouldn't trigger">
+    </div>
+    <div class="form-group">
+      <label for="inputCategory">Category (optional)</label>
+      <input type="text" id="inputCategory" placeholder="e.g. happy-path, edge-case, adversarial">
+    </div>
+    <div class="modal-footer">
+      <button class="btn" onclick="closeModal()">Cancel</button>
+      <button class="btn btn-primary" onclick="saveEntry()">Save</button>
+    </div>
+  </div>
+</div>
+
+<script>
+(function(){
+  "use strict";
+
+  // --- Data ---
+  var rawData = __EVAL_DATA_PLACEHOLDER__;
+  var evalSet = Array.isArray(rawData) ? rawData : (rawData && rawData.queries ? rawData.queries : []);
+  var currentFilter = "all";
+  var editIndex = -1;
+
+  // --- Render ---
+  function render(){
+    renderStats();
+    renderCards();
+    updateCounts();
+  }
+
+  function renderStats(){
+    var total = evalSet.length;
+    var triggers = evalSet.filter(function(q){return q.should_trigger;}).length;
+    var noTriggers = total - triggers;
+    document.getElementById("stats").innerHTML =
+      '<div class="stat stat-total"><div class="stat-label">Total</div><div class="stat-value">'+total+'</div></div>'+
+      '<div class="stat stat-trigger"><div class="stat-label">Should Trigger</div><div class="stat-value">'+triggers+'</div></div>'+
+      '<div class="stat stat-no-trigger"><div class="stat-label">Should Not</div><div class="stat-value">'+noTriggers+'</div></div>';
+  }
+
+  function renderCards(){
+    var container = document.getElementById("cardContainer");
+    var filtered = evalSet.filter(function(q, i){
+      q._index = i;
+      if(currentFilter === "trigger") return q.should_trigger;
+      if(currentFilter === "no-trigger") return !q.should_trigger;
+      return true;
+    });
+
+    if(filtered.length === 0){
+      container.innerHTML = '<div class="empty">No queries to display. Click "+ Add Query" to create one.</div>';
+      return;
+    }
+
+    container.innerHTML = filtered.map(function(q){
+      var cls = q.should_trigger ? "trigger" : "no-trigger";
+      var badge = q.should_trigger
+        ? '<span class="card-badge badge-trigger">Should Trigger</span>'
+        : '<span class="card-badge badge-no-trigger">Should Not Trigger</span>';
+      var meta = '';
+      if(q.category) meta += '<span>Category: '+esc(q.category)+'</span>';
+      if(q.confidence) meta += '<span>Confidence: '+q.confidence+'</span>';
+      var reason = q.reason ? '<div class="card-reason">'+esc(q.reason)+'</div>' : '';
+      return '<div class="card '+cls+'">'+
+        '<div class="card-header"><div class="card-query">'+esc(q.query)+'</div>'+badge+'</div>'+
+        (meta ? '<div class="card-meta">'+meta+'</div>' : '')+
+        reason+
+        '<div class="card-actions">'+
+          '<button class="btn" onclick="openModal(\'edit\','+q._index+')">Edit</button>'+
+          '<button class="btn btn-danger" onclick="deleteEntry('+q._index+')">Delete</button>'+
+        '</div></div>';
+    }).join("");
+  }
+
+  function updateCounts(){
+    var total = evalSet.length;
+    var triggers = evalSet.filter(function(q){return q.should_trigger;}).length;
+    document.getElementById("countAll").textContent = total;
+    document.getElementById("countTrigger").textContent = triggers;
+    document.getElementById("countNoTrigger").textContent = total - triggers;
+  }
+
+  function esc(s){
+    if(!s) return "";
+    var d = document.createElement("div");
+    d.textContent = s;
+    return d.innerHTML;
+  }
+
+  // --- Filters ---
+  document.querySelectorAll(".filter-btn").forEach(function(btn){
+    btn.addEventListener("click", function(){
+      document.querySelectorAll(".filter-btn").forEach(function(b){b.classList.remove("active");});
+      btn.classList.add("active");
+      currentFilter = btn.getAttribute("data-filter");
+      renderCards();
+    });
+  });
+
+  // --- Modal ---
+  window.openModal = function(mode, idx){
+    editIndex = (mode === "edit") ? idx : -1;
+    document.getElementById("modalTitle").textContent = mode === "edit" ? "Edit Query" : "Add Query";
+    if(mode === "edit" && idx >= 0){
+      var q = evalSet[idx];
+      document.getElementById("inputQuery").value = q.query || "";
+      document.getElementById("inputTrigger").value = q.should_trigger ? "true" : "false";
+      document.getElementById("inputReason").value = q.reason || "";
+      document.getElementById("inputCategory").value = q.category || "";
+    } else {
+      document.getElementById("inputQuery").value = "";
+      document.getElementById("inputTrigger").value = "true";
+      document.getElementById("inputReason").value = "";
+      document.getElementById("inputCategory").value = "";
+    }
+    document.getElementById("modalOverlay").classList.add("open");
+    document.getElementById("inputQuery").focus();
+  };
+
+  window.closeModal = function(){
+    document.getElementById("modalOverlay").classList.remove("open");
+    editIndex = -1;
+  };
+
+  window.saveEntry = function(){
+    var query = document.getElementById("inputQuery").value.trim();
+    if(!query){ alert("Query is required."); return; }
+    var entry = {
+      query: query,
+      should_trigger: document.getElementById("inputTrigger").value === "true"
+    };
+    var reason = document.getElementById("inputReason").value.trim();
+    var category = document.getElementById("inputCategory").value.trim();
+    if(reason) entry.reason = reason;
+    if(category) entry.category = category;
+
+    if(editIndex >= 0){
+      evalSet[editIndex] = entry;
+    } else {
+      evalSet.push(entry);
+    }
+    closeModal();
+    render();
+  };
+
+  window.deleteEntry = function(idx){
+    if(!confirm("Delete this query?")) return;
+    evalSet.splice(idx, 1);
+    render();
+  };
+
+  // --- Download ---
+  window.downloadJSON = function(){
+    var out = {
+      skill_name: document.getElementById("skillName").textContent,
+      description: document.getElementById("skillDesc").textContent,
+      queries: evalSet
+    };
+    var blob = new Blob([JSON.stringify(out, null, 2)], {type:"application/json"});
+    var a = document.createElement("a");
+    a.href = URL.createObjectURL(blob);
+    a.download = "eval_set.json";
+    a.click();
+    URL.revokeObjectURL(a.href);
+  };
+
+  // --- Keyboard ---
+  document.addEventListener("keydown", function(e){
+    if(e.key === "Escape") closeModal();
+  });
+
+  // --- Init ---
+  render();
+})();
+</script>
+</body>
+</html>

package/.ai-rules/skills/skill-creator/assets/skill-template.md

@@ -0,0 +1,43 @@
+---
+name: {{SKILL_NAME}}
+description: TODO - describe when to use this skill with specific trigger phrases
+---
+
+# {{SKILL_DISPLAY_NAME}}
+
+## Overview
+
+TODO - one sentence describing what this skill does.
+
+**Core principle:** TODO - the single most important rule.
+
+**Iron Law:**
+```
+TODO - the non-negotiable constraint
+```
+
+## When to Use
+
+- TODO - specific situation 1
+- TODO - specific situation 2
+
+**Use this ESPECIALLY when:**
+- TODO
+
+## When NOT to Use
+
+- TODO - what this skill does NOT handle
+
+## The Process
+
+### Phase 1: TODO
+
+TODO - step by step instructions
+
+### Phase 2: TODO
+
+TODO
+
+## Additional resources
+
+- For detailed reference, see [references/TODO.md](references/TODO.md)