skillwiki 0.4.3 → 0.4.4

package/dist/cli.js

@@ -2402,6 +2402,46 @@ async function runDedup(input) {
   };
 }
 
+// src/commands/raw-body-dedup.ts
+import { createHash as createHash2 } from "crypto";
+async function runRawBodyDedup(vault) {
+  const scan = await scanVault(vault);
+  if (!scan.ok) return { exitCode: ExitCode.VAULT_PATH_INVALID, result: scan };
+  const bodyHashMap = /* @__PURE__ */ new Map();
+  let totalFiles = 0;
+  for (const raw of scan.data.raw) {
+    const text = await readPage(raw);
+    const split = splitFrontmatter(text);
+    if (!split.ok) continue;
+    totalFiles++;
+    const bodyHash = createHash2("sha256").update(split.data.body).digest("hex");
+    const fm = extractFrontmatter(text);
+    let fmSha256 = null;
+    if (fm.ok && typeof fm.data.sha256 === "string" && fm.data.sha256.length === 64) {
+      fmSha256 = fm.data.sha256;
+    }
+    const existing = bodyHashMap.get(bodyHash);
+    if (existing) {
+      existing.push({ relPath: raw.relPath, sha256: fmSha256 });
+    } else {
+      bodyHashMap.set(bodyHash, [{ relPath: raw.relPath, sha256: fmSha256 }]);
+    }
+  }
+  const duplicates = [];
+  for (const [bodyHash, files] of bodyHashMap) {
+    if (files.length < 2) continue;
+    const uniqueShas = new Set(files.map((f) => f.sha256));
+    const allHaveSameValidSha = uniqueShas.size === 1 && files.every((f) => f.sha256 !== null);
+    if (!allHaveSameValidSha) {
+      duplicates.push({ bodyHash, files });
+    }
+  }
+  return {
+    exitCode: 0,
+    result: ok({ scanned: totalFiles, duplicates })
+  };
+}
+
 // src/commands/lint.ts
 var STRUCT_MIN_BODY_LINES = 60;
 var STRUCT_MIN_SECTIONS = 3;
@@ -2434,7 +2474,7 @@ function extractSourceEntries(rawFm) {
   return entries;
 }
 var ERROR_ORDER = ["broken_wikilinks", "invalid_frontmatter", "raw_dedup", "broken_sources", "tag_not_in_taxonomy"];
-var WARNING_ORDER = ["index_incomplete", "index_link_format", "stale_page", "page_too_large", "log_rotate_needed", "orphans", "compound_refs", "legacy_citation_style", "orphaned_citations", "duplicate_frontmatter", "work_item_health", "orphaned_project_pages", "missing_overview"];
+var WARNING_ORDER = ["raw_body_duplicate", "raw_subdirectory_duplicate", "index_incomplete", "index_link_format", "stale_page", "page_too_large", "log_rotate_needed", "orphans", "compound_refs", "legacy_citation_style", "orphaned_citations", "duplicate_frontmatter", "work_item_health", "orphaned_project_pages", "missing_overview"];
 var INFO_ORDER = ["bridges", "page_structure", "topic_map_recommended", "frontmatter_wikilink", "wikilink_citation", "missing_tldr", "missing_diagram"];
 async function runLint(input) {
   const buckets = {};
@@ -2484,12 +2524,41 @@ async function runLint(input) {
   }
   const dedup = await runDedup({ vault: input.vault });
   if (dedup.result.ok && dedup.result.data.duplicates.length > 0) buckets.raw_dedup = dedup.result.data.duplicates;
+  const bodyDedup = await runRawBodyDedup(input.vault);
+  if (bodyDedup.result.ok && bodyDedup.result.data.duplicates.length > 0) {
+    buckets.raw_body_duplicate = bodyDedup.result.data.duplicates.map((d) => ({
+      body_hash: d.bodyHash.slice(0, 12),
+      files: d.files.map((f) => `${f.relPath} (sha256: ${f.sha256 ?? "none"})`)
+    }));
+  }
   const compoundRefs = await validateCompoundReferences(input.vault);
   if (compoundRefs.ok && compoundRefs.data.length > 0) buckets.compound_refs = compoundRefs.data;
   const scan = await scanVault(input.vault);
   const allPages = scan.ok ? [...scan.data.typedKnowledge, ...scan.data.raw, ...scan.data.workItems, ...scan.data.compound] : [];
   const slugs = scan.ok ? buildSlugMap(allPages) : /* @__PURE__ */ new Map();
   if (scan.ok) {
+    const subDirDupes = [];
+    const flatStems = /* @__PURE__ */ new Map();
+    const deepFiles = [];
+    for (const raw of scan.data.raw) {
+      const parts = raw.relPath.split("/");
+      if (parts.length === 3) {
+        const stem = parts[2].replace(/\.md$/, "");
+        flatStems.set(`${parts[1]}/${stem}`, raw.relPath);
+      } else if (parts.length > 3) {
+        const stem = parts[parts.length - 1].replace(/\.md$/, "");
+        deepFiles.push({ relPath: raw.relPath, stem, parentType: parts[1] });
+      }
+    }
+    for (const df of deepFiles) {
+      const flatPath = flatStems.get(`${df.parentType}/${df.stem}`);
+      if (flatPath) {
+        subDirDupes.push(`${df.relPath} -> duplicate of ${flatPath}`);
+      }
+    }
+    if (subDirDupes.length > 0) {
+      buckets.raw_subdirectory_duplicate = subDirDupes;
+    }
     const legacyPages = [];
     const orphanedPages = [];
     const structFlags = [];
@@ -3444,18 +3513,6 @@ function checkS3MountPerf(resolvedPath) {
   if (!existsSync6(conceptsDir)) {
     return check("pass", "s3_mount_perf", "S3 mount performance", `S3 FUSE mount (${mountPoint}), no concepts/ to benchmark`);
   }
-  let rgAvailable = false;
-  try {
-    execSync("which rg", { encoding: "utf8", stdio: ["pipe", "pipe", "pipe"] });
-    rgAvailable = true;
-  } catch {
-    return check(
-      "info",
-      "s3_mount_perf",
-      "S3 mount performance",
-      `S3 FUSE mount (${mountPoint}) \u2014 ripgrep not found, benchmark skipped`
-    );
-  }
   const start = Date.now();
   let timedOut = false;
   try {
@@ -3618,7 +3675,7 @@ async function runArchive(input) {
 }
 
 // src/commands/drift.ts
-import { createHash as createHash2 } from "crypto";
+import { createHash as createHash3 } from "crypto";
 import { writeFile as writeFile10 } from "fs/promises";
 
 // src/utils/fetch.ts
@@ -3698,7 +3755,7 @@ async function runDrift(input) {
       });
       continue;
     }
-    const currentHash = createHash2("sha256").update(Buffer.from(resp.data.body, "utf8")).digest("hex");
+    const currentHash = createHash3("sha256").update(Buffer.from(resp.data.body, "utf8")).digest("hex");
     const drifted2 = currentHash !== storedHash;
     if (drifted2 && input.apply) {
       const newFm = rawFrontmatter.replace(/^sha256:\s*[a-f0-9]+$/m, `sha256: ${currentHash}`);
@@ -4683,7 +4740,7 @@ no compound entries found`;
 import { mkdir as mkdir11, writeFile as writeFile15 } from "fs/promises";
 import { existsSync as existsSync9, statSync as statSync3 } from "fs";
 import { join as join29 } from "path";
-import { createHash as createHash3 } from "crypto";
+import { createHash as createHash4 } from "crypto";
 var ALLOWED_KINDS = /* @__PURE__ */ new Set(["note", "bug", "task", "idea", "session-log"]);
 function slugify2(text) {
   const words = text.trim().split(/\s+/).slice(0, 6).join("-").toLowerCase().replace(/[^a-z0-9-]/g, "").replace(/-+/g, "-").replace(/^-|-$/g, "");
@@ -4727,7 +4784,7 @@ async function runObserve(input) {
   const body = `
 ${input.text.trim()}
 `;
-  const sha256 = createHash3("sha256").update(Buffer.from(body, "utf8")).digest("hex");
+  const sha256 = createHash4("sha256").update(Buffer.from(body, "utf8")).digest("hex");
   const frontmatterLines = [
     "---",
     "source_url:",
@@ -4765,7 +4822,7 @@ ${input.text.trim()}
 // src/commands/ingest.ts
 import { readFile as readFile20, writeFile as writeFile16, mkdir as mkdir12 } from "fs/promises";
 import { join as join30 } from "path";
-import { createHash as createHash4 } from "crypto";
+import { createHash as createHash5 } from "crypto";
 var ALLOWED_TYPES = /* @__PURE__ */ new Set(["entity", "concept", "comparison", "query"]);
 var TYPE_DIR = {
   entity: "entities",
@@ -4931,7 +4988,7 @@ async function runIngest(input) {
       };
     }
   }
-  const sha256 = createHash4("sha256").update(Buffer.from(sourceContent, "utf8")).digest("hex");
+  const sha256 = createHash5("sha256").update(Buffer.from(sourceContent, "utf8")).digest("hex");
   const today = todayIso();
   const slug = slugify3(input.title);
   const tags = input.tags && input.tags.length > 0 ? input.tags : [];

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "skillwiki",
-  "version": "0.4.3",
+  "version": "0.4.4",
   "type": "module",
   "bin": {
     "skillwiki": "dist/cli.js"

package/skills/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "skillwiki",
-  "version": "0.4.3",
+  "version": "0.4.4",
   "skills": "./",
   "description": "Project-aware Karpathy-style knowledge base for Claude Code: 18 prompt-only skills (wiki-*, proj-*, using-skillwiki) backed by the deterministic `skillwiki` CLI.",
   "author": {

package/skills/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "skillwiki",
-  "version": "0.4.3",
+  "version": "0.4.4",
   "description": "Project-aware Karpathy-style knowledge base for Codex with 18 prompt-only skills backed by the deterministic skillwiki CLI.",
   "author": {
     "name": "karlorz",

package/skills/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@skillwiki/skills",
-  "version": "0.4.3",
+  "version": "0.4.4",
   "private": true,
   "files": [
     "wiki-*",