@dewtech/dare-cli 3.11.0 → 3.13.0

package/skills/dare-ax/types.ts

@@ -0,0 +1,79 @@
+/**
+ * dare-ax — shared types
+ * License: MIT
+ */
+
+export interface RateLimit {
+  scope: string;
+  limit: number;
+}
+
+export interface Endpoint {
+  method: string;
+  path: string;
+  description?: string;
+}
+
+export interface ProjectConfig {
+  /** Project name */
+  name: string;
+  /** One-paragraph description of what the project does */
+  projectOverview: string;
+  /** Primary language (Rust, TypeScript, Python, Ruby, Go, PHP, etc.) */
+  language: string;
+  /** Framework (Axum, NestJS, FastAPI, Rails, Gin, Laravel, etc.) */
+  framework: string;
+  /** Database (Postgres, MongoDB, Redis, None, etc.) */
+  database: string;
+  /** List of 5-10 key dependencies */
+  keyDependencies: string[];
+  /** 3-5 sentence architecture description */
+  architectureDescription: string;
+  /** ASCII directory tree (optional) */
+  directoryStructure?: string;
+  /** HTTP endpoints (optional) */
+  endpoints?: Endpoint[];
+  /** Config file name (default: "config.json") */
+  configFile?: string;
+  /** Whether project has docker-compose.yml */
+  hasDocker?: boolean;
+  /** Whether project has a Makefile */
+  hasMakefile?: boolean;
+  /** Whether project has tasks.json / Taskfile */
+  hasTaskfile?: boolean;
+  /** Getting started command (default: "make dev") */
+  gettingStartedCommand?: string;
+  /** Rate limit definitions (default: public 100/min, auth 10/min) */
+  rateLimits?: RateLimit[];
+  /** Additional security notes */
+  extraSecurityNotes?: string[];
+  /** CLI binary name (default: project name) */
+  cliBinary?: string;
+  /** Additional notes for AI agents */
+  agentNotes?: string[];
+  /** If true, skip AX validation in CI */
+  axNotApplicable?: boolean;
+}
+
+export interface ValidationResult {
+  valid: boolean;
+  errors: ValidationError[];
+  warnings: ValidationWarning[];
+}
+
+export interface ValidationError {
+  code: string;
+  message: string;
+}
+
+export interface ValidationWarning {
+  code: string;
+  message: string;
+}
+
+export interface MetricResult {
+  id: string;
+  pass: boolean;
+  description: string;
+  detail?: string;
+}

package/skills/dare-ax/validator.ts

@@ -0,0 +1,238 @@
+/**
+ * dare-ax — DareAxValidator
+ * Validates the structure and security of llms.txt.
+ * License: MIT
+ */
+
+import fs from 'fs';
+import { ValidationResult, ValidationError, ValidationWarning } from './types.js';
+import { findAllSecrets } from './secret-detector.js';
+
+/** Required top-level sections in a valid llms.txt */
+const REQUIRED_SECTIONS = [
+  'Project Overview',
+  'Tech Stack',
+  'Architecture',
+  'Key Endpoints',
+  'For AI Agents',
+] as const;
+
+/** Optional but recommended sections */
+const RECOMMENDED_SECTIONS = [
+  'Directory Structure',
+  'Getting Started',
+  'Rate Limits',
+  'Security Notes',
+  'Important Files',
+] as const;
+
+/**
+ * Minimum character count for a non-trivial section body.
+ * A section with < MIN_SECTION_BODY chars is considered empty/stub.
+ */
+const MIN_SECTION_BODY = 10;
+
+export class DareAxValidator {
+  /**
+   * Validates a llms.txt file at the given path.
+   *
+   * @param llmsTxtPath - Absolute path to the llms.txt file.
+   * @returns ValidationResult with errors and warnings.
+   */
+  validate(llmsTxtPath: string): ValidationResult {
+    const errors: ValidationError[] = [];
+    const warnings: ValidationWarning[] = [];
+
+    // Check file existence
+    if (!fs.existsSync(llmsTxtPath)) {
+      errors.push({
+        code: 'E_FILE_NOT_FOUND',
+        message: `llms.txt not found at path: ${llmsTxtPath}`,
+      });
+      return { valid: false, errors, warnings };
+    }
+
+    let content: string;
+    try {
+      content = fs.readFileSync(llmsTxtPath, 'utf-8');
+    } catch (err) {
+      errors.push({
+        code: 'E_FILE_READ_ERROR',
+        message: `Failed to read llms.txt: ${(err as Error).message}`,
+      });
+      return { valid: false, errors, warnings };
+    }
+
+    // Check empty file
+    if (!content.trim()) {
+      errors.push({
+        code: 'E_EMPTY_FILE',
+        message: 'llms.txt is empty.',
+      });
+      return { valid: false, errors, warnings };
+    }
+
+    // Parse sections: lines starting with "## Section Name"
+    const sections = parseSections(content);
+
+    // Check required sections
+    for (const required of REQUIRED_SECTIONS) {
+      if (!sections.has(required)) {
+        errors.push({
+          code: 'E_MISSING_SECTION',
+          message: `Required section "## ${required}" is missing from llms.txt.`,
+        });
+      } else {
+        const body = sections.get(required)!.trim();
+        if (body.length < MIN_SECTION_BODY) {
+          errors.push({
+            code: 'E_EMPTY_SECTION',
+            message: `Section "## ${required}" exists but appears to be empty or a stub (body: "${body.slice(0, 40)}").`,
+          });
+        }
+      }
+    }
+
+    // Check recommended sections (warnings only)
+    for (const recommended of RECOMMENDED_SECTIONS) {
+      if (!sections.has(recommended)) {
+        warnings.push({
+          code: 'W_MISSING_RECOMMENDED_SECTION',
+          message: `Recommended section "## ${recommended}" is missing from llms.txt.`,
+        });
+      }
+    }
+
+    // Check for secrets
+    const secrets = findAllSecrets(content);
+    for (const secret of secrets) {
+      errors.push({
+        code: 'E_SECRET_DETECTED',
+        message: `Secret detected in llms.txt at line ${secret.line}: [${secret.pattern}]. ` +
+          `Content: "${secret.lineContent.slice(0, 80)}". llms.txt is a public file — remove all secrets.`,
+      });
+    }
+
+    // Check that "For AI Agents" section mentions OpenAPI
+    const forAiSection = sections.get('For AI Agents');
+    if (forAiSection && !forAiSection.toLowerCase().includes('openapi')) {
+      warnings.push({
+        code: 'W_MISSING_OPENAPI_REFERENCE',
+        message: 'Section "## For AI Agents" does not mention OpenAPI. Consider adding a reference to /openapi.json.',
+      });
+    }
+
+    // Check that "For AI Agents" section mentions --json
+    if (forAiSection && !forAiSection.includes('--json')) {
+      warnings.push({
+        code: 'W_MISSING_JSON_FLAG_REFERENCE',
+        message: 'Section "## For AI Agents" does not mention --json CLI flag.',
+      });
+    }
+
+    const valid = errors.length === 0;
+    return { valid, errors, warnings };
+  }
+
+  /**
+   * Validates raw llms.txt content (string) without reading from disk.
+   * Useful for previewing generated content before writing.
+   */
+  validateContent(content: string): ValidationResult {
+    const errors: ValidationError[] = [];
+    const warnings: ValidationWarning[] = [];
+
+    if (!content.trim()) {
+      errors.push({ code: 'E_EMPTY_FILE', message: 'Content is empty.' });
+      return { valid: false, errors, warnings };
+    }
+
+    const sections = parseSections(content);
+
+    for (const required of REQUIRED_SECTIONS) {
+      if (!sections.has(required)) {
+        errors.push({
+          code: 'E_MISSING_SECTION',
+          message: `Required section "## ${required}" is missing.`,
+        });
+      } else {
+        const body = sections.get(required)!.trim();
+        if (body.length < MIN_SECTION_BODY) {
+          errors.push({
+            code: 'E_EMPTY_SECTION',
+            message: `Section "## ${required}" is empty or a stub.`,
+          });
+        }
+      }
+    }
+
+    for (const recommended of RECOMMENDED_SECTIONS) {
+      if (!sections.has(recommended)) {
+        warnings.push({
+          code: 'W_MISSING_RECOMMENDED_SECTION',
+          message: `Recommended section "## ${recommended}" is missing.`,
+        });
+      }
+    }
+
+    const secrets = findAllSecrets(content);
+    for (const secret of secrets) {
+      errors.push({
+        code: 'E_SECRET_DETECTED',
+        message: `Secret detected at line ${secret.line}: [${secret.pattern}].`,
+      });
+    }
+
+    const forAiSection = sections.get('For AI Agents');
+    if (forAiSection && !forAiSection.toLowerCase().includes('openapi')) {
+      warnings.push({
+        code: 'W_MISSING_OPENAPI_REFERENCE',
+        message: '"## For AI Agents" does not mention OpenAPI.',
+      });
+    }
+
+    if (forAiSection && !forAiSection.includes('--json')) {
+      warnings.push({
+        code: 'W_MISSING_JSON_FLAG_REFERENCE',
+        message: '"## For AI Agents" does not mention --json flag.',
+      });
+    }
+
+    return { valid: errors.length === 0, errors, warnings };
+  }
+}
+
+/**
+ * Parses section headings (## Section Name) and their body text from a llms.txt string.
+ * Returns a Map<sectionName, bodyText>.
+ */
+function parseSections(content: string): Map<string, string> {
+  const sections = new Map<string, string>();
+  const lines = content.split('\n');
+
+  let currentSection: string | null = null;
+  const bodyLines: string[] = [];
+
+  for (const line of lines) {
+    // Match "## Section Name" (exactly two hashes, not more)
+    const sectionMatch = /^##\s+(.+)$/.exec(line);
+
+    if (sectionMatch) {
+      // Save previous section body
+      if (currentSection !== null) {
+        sections.set(currentSection, bodyLines.join('\n'));
+      }
+      currentSection = sectionMatch[1].trim();
+      bodyLines.length = 0;
+    } else if (currentSection !== null) {
+      bodyLines.push(line);
+    }
+  }
+
+  // Save last section
+  if (currentSection !== null) {
+    sections.set(currentSection, bodyLines.join('\n'));
+  }
+
+  return sections;
+}