@kevinrabun/judges 3.127.1 → 3.127.2

package/README.md

@@ -15,7 +15,7 @@ An MCP (Model Context Protocol) server that provides a panel of **45 specialized
 [![npm](https://img.shields.io/npm/v/@kevinrabun/judges)](https://www.npmjs.com/package/@kevinrabun/judges)
 [![npm downloads](https://img.shields.io/npm/dw/@kevinrabun/judges)](https://www.npmjs.com/package/@kevinrabun/judges)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
-[![Tests](https://img.shields.io/badge/tests-2482-brightgreen)](https://github.com/KevinRabun/judges/actions)
+[![Tests](https://img.shields.io/badge/tests-3614-brightgreen)](https://github.com/KevinRabun/judges/actions)
 
 > 🔰 **Packages**
 > - **CLI**: `@kevinrabun/judges-cli` → binary `judges` (use `npx @kevinrabun/judges-cli eval --file app.ts`).
@@ -843,7 +843,7 @@ The tribunal operates in three layers:
 
 Judges Panel is a **dual-layer** review system: instant **deterministic tools** (offline, no API keys) for pattern and AST analysis, plus **45 expert-persona MCP prompts** that unlock LLM-powered deep analysis when connected to an AI client. It does not try to be a CVE scanner or a linter. Those capabilities belong in dedicated MCP servers that an AI agent can orchestrate alongside Judges.
 
-### Built-in AST Analysis (v2.0.0+)
+### Built-in AST Analysis
 
 Unlike earlier versions that recommended a separate AST MCP server, Judges Panel now includes **real AST-based structural analysis** out of the box:
 
@@ -1236,7 +1236,9 @@ Create a `.judgesrc.json` (or `.judgesrc`) file in your project root to customiz
   "languages": ["typescript", "python"],
   "format": "text",
   "failOnFindings": false,
-  "baseline": ""
+  "baseline": "",
+  "regulatoryScope": ["GDPR", "PCI-DSS", "SOC2"],
+  "consensusThreshold": 0.7
 }
 ```
 
@@ -1252,6 +1254,14 @@ Create a `.judgesrc.json` (or `.judgesrc`) file in your project root to customiz
 | `format` | `string` | `"text"` | Default output format: `text` · `json` · `sarif` · `markdown` · `html` · `pdf` · `junit` · `codeclimate` · `github-actions` |
 | `failOnFindings` | `boolean` | `false` | Exit code 1 when verdict is `fail` — useful for CI gates |
 | `baseline` | `string` | `""` | Path to a baseline JSON file — matching findings are suppressed |
+| `plugins` | `string[]` | `[]` | Plugin module specifiers (npm packages or relative paths) that export custom judges |
+| `judgeWeights` | `object` | `{}` | Weighted importance per judge for aggregated scoring (e.g. `{ "cybersecurity": 2.0 }`) |
+| `failOnScoreBelow` | `number` | — | Minimum score (0–100) for the run to pass; complements `failOnFindings` |
+| `regulatoryScope` | `string[]` | — | Regulatory frameworks in scope (e.g. `["GDPR", "PCI-DSS"]`). Findings citing ONLY out-of-scope frameworks are suppressed. Run `judges list --frameworks` for supported values. |
+| `consensusThreshold` | `number` | — | Consensus suppression (0–1). If this fraction of judges report zero findings, minority findings are suppressed. Recommended: `0.7` for CI. |
+| `escalationThreshold` | `number` | — | Confidence threshold (0–1) below which findings are flagged for human review |
+| `overrides` | `array` | `[]` | Path-scoped config overrides (e.g. `[{ "files": "**/*.test.ts", "disabledJudges": ["documentation"] }]`) |
+| `customRules` | `array` | `[]` | User-defined regex-based rules for business logic validation |
 
 All evaluation tools (CLI and MCP) accept the same configuration fields via `--config <path>` or inline `config` parameter.
 
@@ -1288,6 +1298,38 @@ Patches include `oldText`, `newText`, `startLine`, and `endLine` for automated a
 
 When multiple judges flag the same issue (e.g., both Data Security and Cybersecurity detect SQL injection on line 15), findings are automatically deduplicated. The highest-severity finding wins, and the description is annotated with cross-references (e.g., *"Also identified by: CYBER-003"*).
 
+### Human Focus Guide
+
+Every tribunal evaluation includes a `humanFocusGuide` that categorizes findings into three buckets for human reviewers:
+
+| Bucket | Description | When to use |
+|--------|-------------|-------------|
+| **✅ Trust** | High-confidence (≥80%), evidence-backed findings with AST/taint confirmation | Act directly — these have strong automated evidence |
+| **🔍 Verify** | Lower-confidence or absence-based findings | Use your judgment — the issue may exist elsewhere in the project |
+| **🔦 Blind Spots** | Areas automated analysis cannot evaluate | Focus your manual review time here |
+
+Blind spots are detected from code characteristics: complex branching logic, external service calls, financial calculations, PII handling, state machines, and complex regex. The guide appears in CLI text/markdown output, JSON/SARIF output, and GitHub Action step summaries.
+
+### Regulatory Scope
+
+Configure which regulatory frameworks apply to your project in `.judgesrc`:
+
+```json
+{ "regulatoryScope": ["GDPR", "PCI-DSS", "SOC2"] }
+```
+
+Findings that cite ONLY out-of-scope frameworks are suppressed. Findings with no regulatory reference (general code quality) are always kept. Run `judges list --frameworks` to see all 17 supported frameworks (GDPR, CCPA, HIPAA, PCI-DSS, SOC2, SOX, COPPA, FedRAMP, NIST, ISO27001, ePrivacy, DORA, NIS2, EU-AI-Act, and more).
+
+### Self-Teaching Amendments
+
+The LLM benchmark system auto-generates precision amendments for judges with high false-positive rates. Amendments are data-driven corrections injected into prompts that improve accuracy over successive benchmark runs.
+
+The self-teaching loop:
+1. Run benchmark → analyzer identifies judges below 70% precision
+2. Generates targeted amendments (e.g., "Judge ERR: do not flag clean Express code with framework error middleware")
+3. Next benchmark run loads amendments → precision improves
+4. Run `judges codify-amendments` to bake amendments permanently into the distributed package
+
 ### Taint Flow Analysis
 
 The engine performs inter-procedural taint tracking to trace data from user-controlled sources (e.g., `req.body`, `process.env`) through transformations to security-sensitive sinks (e.g., `eval()`, `exec()`, SQL queries). Taint flows are used to boost confidence on true-positive findings and suppress false positives where sanitization is detected.
@@ -1475,6 +1517,8 @@ judges/
 | `judges config import <src>` | Import a shared configuration |
 | `judges compare` | Compare judges against other code review tools |
 | `judges list` | List all 45 judges with domains and descriptions |
+| `judges list --frameworks` | List supported regulatory frameworks and `.judgesrc` usage |
+| `judges codify-amendments` | Bake self-teaching amendments into judge source files |
 
 ---
 

package/dist/commands/llm-benchmark.js

@@ -153,7 +153,7 @@ export function parseLlmRuleIds(response) {
     // IDs mentioned in rationale text or findings tables of "clean" judge sections
     // from being counted as detections.
     const sections = response.split(/(?:^|\n)---\s*\n|(?=^## )/m);
-    const zeroFindingsPattern = /\*?\*?(?:ZERO|zero|0|no)\s+findings?\*?\*?|(?:findings?|issues?)[\s:]*\*?\*?(?:none|0|zero)\*?\*?|no\s+(?:issues?|findings?|problems?|concerns?)\s+(?:found|detected|identified|reported)|report(?:ing)?\s+zero|Score\s*[|:]\s*\*?\*?100\s*\/?\s*100\*?\*?/i;
+    const zeroFindingsPattern = /(?:ZERO|zero|0|no) findings?|findings?[:\s]*(?:none|0|zero)|no (?:issues|findings|problems|concerns) (?:found|detected|identified|reported)|reporting? zero|Score[|: ]*100/i;
     for (const section of sections) {
         // If this section explicitly declares zero/no findings or a perfect score,
         // skip rule ID extraction — any rule IDs are explanatory references

package/dist/evaluators/index.js

@@ -504,7 +504,7 @@ function synthesizeHumanFocusGuide(findings, code, language) {
             });
         }
         // State machines / workflow
-        const hasStateMachine = /state\s*[=:]\s*['"][^'"]+['"]|status\s*===?\s*['"]|transition|workflow|step.*next/i.test(code);
+        const hasStateMachine = /state\s*[=:]\s*['"][^'"]+['"]|status\s*===?\s*['"]|transition|workflow|step[\w\s]{0,20}next/i.test(code);
         if (hasStateMachine) {
             blindSpots.push({
                 area: "State Management / Workflow Logic",

package/dist/probabilistic/llm-response-validator.js

@@ -5,7 +5,10 @@ const SEVERITY_SET = new Set(["critical", "high", "medium", "low", "info"]);
  * Attempt to parse a JSON payload embedded in LLM output. Supports fenced code blocks and raw JSON.
  */
 function parseJsonBlock(text) {
-    const fenceMatch = text.match(/```(?:json)?[ \t]*\n([\s\S]*?)\n[ \t]*```/i) ?? text.match(/```(?:json)?[ \t]*([\s\S]*?)```/i);
+    // Extract JSON from fenced code blocks — limit search to first 50KB to prevent ReDoS on large input
+    const searchText = text.length > 50_000 ? text.slice(0, 50_000) : text;
+    const fenceMatch = searchText.match(/```(?:json)?\s*\n([\s\S]{0,20000}?)\n\s*```/i) ??
+        searchText.match(/```(?:json)?\s*([\s\S]{0,20000}?)```/i);
     if (fenceMatch) {
         try {
             return JSON.parse(fenceMatch[1]);

package/dist/reports/public-repo-report.js

@@ -216,13 +216,9 @@ function compileExcludeRegexes(patterns) {
     if (!patterns || patterns.length === 0)
         return [];
     return patterns.map((pattern) => {
-        try {
-            return new RegExp(pattern, "i");
-        }
-        catch {
-            // Invalid regex from user input — treat as literal string match
-            return new RegExp(pattern.replace(/[.*+?^${}()|[\]\\]/g, "\\$&"), "i");
-        }
+        // Always escape user input to prevent regex injection, then compile
+        const escaped = pattern.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
+        return new RegExp(escaped, "i");
     });
 }
 function isLikelyNonProductionPath(path) {

package/dist/skill-loader.js

@@ -25,7 +25,7 @@ export function parseSkillFrontmatter(raw) {
             i++;
             continue;
         }
-        const kv = line.match(/^([a-zA-Z_][a-zA-Z0-9_-]*)[ \t]*:[ \t]*(.*)$/);
+        const kv = line.match(/^([a-zA-Z_][a-zA-Z0-9_-]*)[ \t]*:[ \t]*(.*?)$/s);
         if (!kv) {
             i++;
             continue;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@kevinrabun/judges",
-  "version": "3.127.1",
+  "version": "3.127.2",
   "description": "45 specialized judges that evaluate AI-generated code for security, cost, and quality.",
   "mcpName": "io.github.KevinRabun/judges",
   "type": "module",

package/packages/judges-cli/README.md

@@ -1,6 +1,6 @@
 # @kevinrabun/judges-cli
 
-Standalone CLI package for Judges.
+Standalone CLI package for the [Judges Panel](https://github.com/KevinRabun/judges) — 45 specialized judges that evaluate code for security, quality, compliance, and 40 more dimensions.
 
 ## Install
 
@@ -11,14 +11,46 @@ npm install -g @kevinrabun/judges-cli
 ## Usage
 
 ```bash
+# Evaluate code
 judges eval src/app.ts
+judges eval src/ --format sarif --output report.sarif
+judges eval src/app.ts --judge cybersecurity
+judges eval src/app.ts --preset strict --fail-on-findings
+
+# List judges and regulatory frameworks
 judges list
-judges hook install
+judges list --frameworks
+
+# Auto-fix findings
+judges fix src/app.ts --apply
 
 # Agentic skills
 judges skill ai-code-review --file src/app.ts
 judges skill security-review --file src/api.ts --format json
-judges skills   # list available skills
+judges skills
+
+# Self-teaching
+judges codify-amendments          # bake benchmark amendments into judge files
+judges codify-amendments --dry-run
 ```
 
-Use `@kevinrabun/judges` when you need the MCP server or programmatic API.
+## Configuration
+
+Create a `.judgesrc.json` in your project root:
+
+```json
+{
+  "preset": "strict",
+  "regulatoryScope": ["GDPR", "PCI-DSS"],
+  "disabledJudges": ["accessibility"],
+  "failOnFindings": true
+}
+```
+
+See the [full configuration reference](https://github.com/KevinRabun/judges#configuration) for all options.
+
+## Packages
+
+- **`@kevinrabun/judges-cli`** — This package. Binary `judges` for CI/CD pipelines.
+- **`@kevinrabun/judges`** — Programmatic API + MCP server.
+- **VS Code extension** — [`kevinrabun.judges-panel`](https://marketplace.visualstudio.com/items?itemName=kevinrabun.judges-panel).

package/server.json

@@ -16,12 +16,12 @@
       "mimeType": "image/png"
     }
   ],
-  "version": "3.127.1",
+  "version": "3.127.2",
   "packages": [
     {
       "registryType": "npm",
       "identifier": "@kevinrabun/judges",
-      "version": "3.127.1",
+      "version": "3.127.2",
       "transport": {
         "type": "stdio"
       }

package/src/skill-loader.ts

@@ -44,7 +44,7 @@ export function parseSkillFrontmatter(raw: string): { meta: SkillMeta; body: str
       i++;
       continue;
     }
-    const kv = line.match(/^([a-zA-Z_][a-zA-Z0-9_-]*)[ \t]*:[ \t]*(.*)$/);
+    const kv = line.match(/^([a-zA-Z_][a-zA-Z0-9_-]*)[ \t]*:[ \t]*(.*?)$/s);
     if (!kv) {
       i++;
       continue;