feat(sync-gbrain): default to --strategy auto (code + markdown)

The /sync-gbrain skill previously ran 'gbrain sync --strategy code' for
the code-stage walk, silently skipping every markdown file in the repo. For
code repos with non-trivial docs (READMEs aside — those are still skipped
by gbrain's own hard-coded list), this meant guides, ADRs, design docs,
and architectural notes never reached the brain.

Switch the page-creating walk in the code stage to --strategy auto so a
single sync pass indexes both code AND markdown. Full-mode behavior is
preserved: after the walk, --full still runs 'gbrain reindex-code --yes'
to force chunker-version re-embed of code pages (upstream garrytan#1639/garrytan#1642's
walk+reindex split, kept intact).

Three call sites updated: the dry-run summary string, the spawnGbrain
walk call, and the walk's error-summary message. Doc block at top of
file and the inline comment block above the walk are updated to describe
the new behavior. sync-gbrain/SKILL.md.tmpl's architecture note
('native code surfaces ... gbrain sync --strategy auto') is updated to
match; generated SKILL.md files regenerated via 'bun run gen:skill-docs
--host all' (all 9 hosts + Claude default).

Personal fork only — not pushed upstream.

Author: nobody

bin/gstack-gbrain-sync.ts

@@ -4,11 +4,15 @@
  *
  * Orchestrates three storage tiers per plan §"Storage tiering":
  *
- *   1. Code (current repo)         → `gbrain sources add` (idempotent via
- *                                    lib/gbrain-sources.ts) + `gbrain sync
- *                                    --strategy code` (incremental) or
- *                                    `gbrain reindex-code --yes` (--full).
- *                                    NEVER `gbrain import` (markdown only).
+ *   1. Code + markdown (current repo) → `gbrain sources add` (idempotent
+ *                                    via lib/gbrain-sources.ts) + `gbrain
+ *                                    sync --strategy auto` (incremental;
+ *                                    walks code AND markdown in one pass).
+ *                                    --full additionally runs `gbrain
+ *                                    reindex-code --yes` after the walk
+ *                                    to force chunker-version re-embed of
+ *                                    code pages. NEVER `gbrain import`
+ *                                    (transcript pipeline).
  *   2. Transcripts + curated memory → gstack-memory-ingest (typed put_page)
  *   3. Curated artifacts to git    → gstack-brain-sync (existing pipeline)
  *
@@ -631,7 +635,7 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
       ran: false,
       ok: true,
       duration_ms: 0,
-      summary: `would: gbrain sources add ${sourceId} --path ${root} --federated; gbrain sync --strategy code --source ${sourceId}; gbrain sources attach ${sourceId}`,
+      summary: `would: gbrain sources add ${sourceId} --path ${root} --federated; gbrain sync --strategy auto --source ${sourceId}; gbrain sources attach ${sourceId}`,
       detail: { source_id: sourceId, source_path: root, status: "skipped" },
     };
   }
@@ -713,14 +717,15 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
   // run that called reindex-code alone found nothing ("No code pages to
   // reindex"), finished in ~1s, and left the code index permanently empty
   // while still reporting OK. The page-creating walk is `sync --strategy
-  // code`, so --full must run it FIRST, then reindex-code, to honor the
-  // documented "full walk + reindex" contract for both fresh and populated
-  // sources.
+  // auto` (which walks code AND markdown — formerly `--strategy code`, now
+  // `auto` so we index repo docs too), so --full must run it FIRST, then
+  // reindex-code, to honor the documented "full walk + reindex" contract
+  // for both fresh and populated sources.
   const codeTimeoutMs = resolveStageTimeoutMs(
     process.env.GSTACK_SYNC_CODE_TIMEOUT_MS,
     "GSTACK_SYNC_CODE_TIMEOUT_MS",
   );
-  const walkResult = spawnGbrain(["sync", "--strategy", "code", "--source", sourceId], {
+  const walkResult = spawnGbrain(["sync", "--strategy", "auto", "--source", sourceId], {
     stdio: args.quiet ? ["ignore", "ignore", "ignore"] : ["ignore", "inherit", "inherit"],
     timeout: codeTimeoutMs,
     baseEnv: gbrainEnv,
@@ -732,7 +737,7 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
       ran: true,
       ok: false,
       duration_ms: Date.now() - t0,
-      summary: `gbrain sync --strategy code --source ${sourceId} exited ${walkResult.status}`,
+      summary: `gbrain sync --strategy auto --source ${sourceId} exited ${walkResult.status}`,
       detail: { source_id: sourceId, source_path: root, status: "failed" },
     };
   }

sync-gbrain/SKILL.md

@@ -755,7 +755,7 @@ guidance in CLAUDE.md so the coding agent knows when to prefer `gbrain`
 search over Grep.
 
 **Architecture (post-codex review):** This skill uses gbrain v0.20.0+'s
-**native code surfaces** (`gbrain sources add`, `gbrain sync --strategy code`,
+**native code surfaces** (`gbrain sources add`, `gbrain sync --strategy auto`,
 `gbrain reindex-code`, `gbrain code-def/code-refs/code-callers/code-callees`).
 It does NOT use `gbrain import` (that path is for markdown directories).
 It does NOT touch `~/.gstack/` indexing (the existing `gstack-gbrain-source-wireup`

sync-gbrain/SKILL.md.tmpl

@@ -35,7 +35,7 @@ guidance in CLAUDE.md so the coding agent knows when to prefer `gbrain`
 search over Grep.
 
 **Architecture (post-codex review):** This skill uses gbrain v0.20.0+'s
-**native code surfaces** (`gbrain sources add`, `gbrain sync --strategy code`,
+**native code surfaces** (`gbrain sources add`, `gbrain sync --strategy auto`,
 `gbrain reindex-code`, `gbrain code-def/code-refs/code-callers/code-callees`).
 It does NOT use `gbrain import` (that path is for markdown directories).
 It does NOT touch `~/.gstack/` indexing (the existing `gstack-gbrain-source-wireup`