fix(cli): report unsupported-language projects instead of finishing silently (#1806)

A project CodeGraph has no grammar for was indistinguishable from an empty one: unsupported extensions are filtered out at discovery, so filesDiscovered was 0, the reconciliation in index.ts found no shortfall and recorded index_state as complete, and the CLI printed the same 'No files found to index' it prints for an empty repo.

That silence is what makes it costly over MCP: an empty result reads identically to 'no match', and the agent has been told to trust the graph rather than grep.

The scan already visits every file, so the tally of what it declined to index costs no extra I/O and no second pass. index_state itself is left alone: changing its values would change the status --json contract, which is a call for the maintainer to make.


(cherry picked from commit 2c20892789897f502f84152f78b777b11d65fdaa)

Co-authored-by: Max Hsu <maxmilian@gmail.com>
Co-authored-by: netbrah <netbrah@users.noreply.github.com>

src/bin/codegraph.ts

@@ -388,6 +388,8 @@ type IndexResult = {
   edgesCreated: number;
   errors: Array<{ message: string; filePath?: string; severity: string; code?: string }>;
   durationMs: number;
+  filesSkippedUnsupported?: number;
+  topUnsupportedExtensions?: { ext: string; count: number }[];
 };
 
 /**
@@ -445,6 +447,20 @@ function printIndexResult(clack: typeof import('@clack/prompts'), result: IndexR
     }
   } else if (hasErrors) {
     clack.log.error(`Indexing failed ${getGlyphs().dash} all ${formatNumber(result.filesErrored)} files had errors`);
+  } else if (result.filesSkippedUnsupported) {
+    // A project CodeGraph has no grammar for used to be indistinguishable from
+    // an empty one: same message, same `complete` state, same exit 0. Say which
+    // files were there and that the graph is empty on purpose, so nobody — and
+    // no agent trusting the graph — reads silence as "this code doesn't exist"
+    // (#1502).
+    const top = (result.topUnsupportedExtensions ?? [])
+      .map(e => `${e.ext} (${formatNumber(e.count)})`)
+      .join(', ');
+    clack.log.warn(
+      `No supported source files found ${getGlyphs().dash} ${formatNumber(result.filesSkippedUnsupported)} file(s) present, none in a language CodeGraph indexes`
+      + (top ? `: ${top}` : '')
+    );
+    clack.log.info('CodeGraph is inactive for this workspace — searches will return nothing. Use your own file tools here.');
   } else {
     clack.log.warn('No files found to index');
   }

src/extraction/index.ts

@@ -99,6 +99,16 @@ export interface IndexResult {
    * counts. Only set by full-index runs (indexAll), not indexFiles/sync.
    */
   filesDiscovered?: number;
+  /**
+   * Files the scan saw but has no grammar for, tallied by extension. Only the
+   * degenerate case needs it: a project of unsupported files otherwise looks
+   * exactly like an empty one (0 files, state `complete`), so nothing tells the
+   * user — or an agent — that there was code here CodeGraph could not read
+   * (#1502). Counted during the scan's existing walk.
+   */
+  filesSkippedUnsupported?: number;
+  /** The most common unsupported extensions, biggest first. */
+  topUnsupportedExtensions?: { ext: string; count: number }[];
   nodesCreated: number;
   edgesCreated: number;
   errors: ExtractionError[];
@@ -1408,9 +1418,30 @@ export function scanDirectory(
  * Async variant of scanDirectory that yields to the event loop periodically,
  * allowing worker threads to receive and render progress messages.
  */
+/**
+ * What a scan saw but could not index, tallied by extension.
+ *
+ * Filled during the walk the scan already performs — a project of unsupported
+ * files is otherwise indistinguishable from an empty one, because unsupported
+ * extensions are filtered out at discovery and never counted anywhere (#1502).
+ */
+export interface ScanSkipStats {
+  /** Lowercased extension (with dot) → how many files carried it. */
+  unsupportedByExtension: Map<string, number>;
+}
+
+/** Record one file the scan declined to index. */
+function tallySkip(stats: ScanSkipStats | undefined, rel: string): void {
+  if (!stats) return;
+  const ext = path.extname(rel).toLowerCase();
+  if (!ext) return;
+  stats.unsupportedByExtension.set(ext, (stats.unsupportedByExtension.get(ext) ?? 0) + 1);
+}
+
 export async function scanDirectoryAsync(
   rootDir: string,
-  onProgress?: (current: number, file: string) => void
+  onProgress?: (current: number, file: string) => void,
+  stats?: ScanSkipStats
 ): Promise<string[]> {
   // Custom extension → language overrides from the project's codegraph.json.
   const overrides = loadExtensionOverrides(rootDir);
@@ -1428,12 +1459,14 @@ export async function scanDirectoryAsync(
         if (count % 100 === 0) {
           await new Promise<void>(r => setImmediate(r));
         }
+      } else {
+        tallySkip(stats, filePath);
       }
     }
     return files;
   }
 
-  return scanDirectoryWalk(rootDir, onProgress);
+  return scanDirectoryWalk(rootDir, onProgress, stats);
 }
 
 /**
@@ -1441,7 +1474,8 @@ export async function scanDirectoryAsync(
  */
 function scanDirectoryWalk(
   rootDir: string,
-  onProgress?: (current: number, file: string) => void
+  onProgress?: (current: number, file: string) => void,
+  stats?: ScanSkipStats
 ): string[] {
   const files: string[] = [];
   let count = 0;
@@ -1524,10 +1558,14 @@ function scanDirectoryWalk(
               walk(fullPath, active);
             }
           } else if (stat.isFile()) {
-            if (!isIgnored(fullPath, false, active) && isSourceFile(relativePath, overrides)) {
-              files.push(relativePath);
-              count++;
-              onProgress?.(count, relativePath);
+            if (!isIgnored(fullPath, false, active)) {
+              if (isSourceFile(relativePath, overrides)) {
+                files.push(relativePath);
+                count++;
+                onProgress?.(count, relativePath);
+              } else {
+                tallySkip(stats, relativePath);
+              }
             }
           }
         } catch {
@@ -1541,10 +1579,14 @@ function scanDirectoryWalk(
           walk(fullPath, active);
         }
       } else if (entry.isFile()) {
-        if (!isIgnored(fullPath, false, active) && isSourceFile(relativePath, overrides)) {
-          files.push(relativePath);
-          count++;
-          onProgress?.(count, relativePath);
+        if (!isIgnored(fullPath, false, active)) {
+          if (isSourceFile(relativePath, overrides)) {
+            files.push(relativePath);
+            count++;
+            onProgress?.(count, relativePath);
+          } else {
+            tallySkip(stats, relativePath);
+          }
         }
       }
     }
@@ -1791,6 +1833,7 @@ export class ExtractionOrchestrator {
     // early-run 5-10s single stalls were observed on 95k-file repos but never
     // attributed — these labels settle scan vs framework-detect vs grammars.
     const tScan = Date.now();
+    const skipStats: ScanSkipStats = { unsupportedByExtension: new Map() };
     const files = await scanDirectoryAsync(this.rootDir, (current, file) => {
       onProgress?.({
         phase: 'scanning',
@@ -1798,8 +1841,20 @@ export class ExtractionOrchestrator {
         total: 0,
         currentFile: file,
       });
-    });
+    }, skipStats);
     if (process.env.CODEGRAPH_SYNTH_TIMINGS) console.error(`[phase-timing] scan: ${Date.now() - tScan}ms (${files.length} files)`);
+    /** Only meaningful when nothing was indexable — see IndexResult (#1502). */
+    const skipSummary = (): Pick<IndexResult, 'filesSkippedUnsupported' | 'topUnsupportedExtensions'> => {
+      let total = 0;
+      for (const n of skipStats.unsupportedByExtension.values()) total += n;
+      if (total === 0) return {};
+      const top = [...skipStats.unsupportedByExtension.entries()]
+        .map(([ext, count]) => ({ ext, count }))
+        .sort((a, b) => b.count - a.count || a.ext.localeCompare(b.ext))
+        .slice(0, 5);
+      return { filesSkippedUnsupported: total, topUnsupportedExtensions: top };
+    };
+
 
     // A re-index over an existing DB skips unchanged-hash files at the store,
     // which would preserve wiped zero-node rows (#1541) — drop them first so
@@ -2191,6 +2246,7 @@ export class ExtractionOrchestrator {
         filesSkipped,
         filesErrored,
         filesDiscovered: total,
+        ...skipSummary(),
         nodesCreated: totalNodes,
         edgesCreated: totalEdges,
         errors: [{ message: 'Aborted', severity: 'error' }, ...errors],
@@ -2347,6 +2403,7 @@ export class ExtractionOrchestrator {
       filesSkipped,
       filesErrored,
       filesDiscovered: total,
+      ...skipSummary(),
       nodesCreated: totalNodes,
       edgesCreated: totalEdges,
       errors,