fix: /sync-gbrain respects an existing valid .gbrain-source pin (#2417)

/sync-gbrain always derived a new worktree-scoped source ID, even when
the repository already carried a valid .gbrain-source pin created through
the native GBrain source workflow — silently bypassing the selected
source boundary, registering a duplicate federated source, and routing
later dream/cycle checks to the wrong source.

Now a local pin is reused when it passes the fail-closed identity checks:
the ID is syntactically valid, the source is registered, and the
registered path realpath-resolves to the current checkout (so a stale or
copied dotfile can't redirect a sync into another repo's source). A
confirmed pin is treated as user-managed — synced and attached without
add/remove, legacy migration, or federation changes. Dry-run stays
spawn-free (reads only the local marker for previews). Missing, invalid,
stale, or unreadable pins fall back to the existing generated source ID.

Absorbs PR #2417 by @exGeni (applied via git am -3; 42 tests pass in
test/gstack-gbrain-sync.test.ts including the new pin-respecting
coverage: spawn-free dry-run, symlink-equivalent registered paths,
non-dry-run sync/attach with no add/remove, dream routing, unreadable
markers, config-backed env use).

Co-authored-by: Evgenii Lopatin <e75533@gmail.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
Garry Tan
2026-08-16 10:02:31 -07:00
co-authored by Evgenii Lopatin Claude Fable 5
parent acc354fcfa
commit bfa579d4ea
2 changed files with 215 additions and 24 deletions
+73 -23
View File
@@ -29,7 +29,7 @@
* than building a gstack-side daemon.
*/
import { existsSync, statSync, mkdirSync, writeFileSync, readFileSync, unlinkSync, renameSync } from "fs";
import { existsSync, statSync, mkdirSync, writeFileSync, readFileSync, unlinkSync, renameSync, realpathSync } from "fs";
import { join, dirname } from "path";
import { execSync, spawnSync } from "child_process";
import { homedir, hostname } from "os";
@@ -368,6 +368,42 @@ function deriveCodeSourceId(repoPath: string): string {
return constrainSourceId("gstack-code", `${base}-${hostPathHash}`);
}
/**
* Reuse an explicit repo pin when it names a registered source for this exact
* checkout. The path check prevents a stale or copied dotfile from redirecting
* a code sync into another repo's source.
*/
function readPinnedSourceId(repoPath: string): string | null {
const pinPath = join(repoPath, ".gbrain-source");
if (!existsSync(pinPath)) return null;
try {
const sourceId = readFileSync(pinPath, "utf-8").trim();
return /^[a-z0-9](?:[a-z0-9-]{0,30}[a-z0-9])?$/.test(sourceId) ? sourceId : null;
} catch {
// A pin is advisory. A permission race or a directory at this path must
// not turn a sync preview into an unexpected crash.
return null;
}
}
export function existingPinnedSourceId(repoPath: string, env?: NodeJS.ProcessEnv): string | null {
const sourceId = readPinnedSourceId(repoPath);
if (!sourceId) return null;
const registeredPath = sourceLocalPath(sourceId, env);
if (!registeredPath) return null;
try {
return realpathSync(registeredPath) === realpathSync(repoPath) ? sourceId : null;
} catch {
return null;
}
}
function resolveCodeSourceId(repoPath: string, env?: NodeJS.ProcessEnv): string {
return existingPinnedSourceId(repoPath, env) ?? deriveCodeSourceId(repoPath);
}
/**
* Pre-pathhash source id, kept for orphan detection only.
*
@@ -820,7 +856,13 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
return { name: "code", ran: false, ok: true, duration_ms: 0, summary: "skipped (not in git repo)" };
}
const sourceId = deriveCodeSourceId(root);
// A preview must not spawn gbrain. Trust a syntactically-valid local pin
// there; a real run confirms its registered path before using it.
const gbrainEnv = args.mode === "dry-run" ? undefined : buildGbrainEnv({ announce: !args.quiet });
const pinnedSourceId = args.mode === "dry-run"
? readPinnedSourceId(root)
: existingPinnedSourceId(root, gbrainEnv);
const sourceId = pinnedSourceId ?? deriveCodeSourceId(root);
// Per-repo trust tier — checked BEFORE the dry-run branch so previews report
// the refusal honestly instead of claiming they would sync.
@@ -861,7 +903,9 @@ 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: pinnedSourceId
? `would: gbrain sync --strategy code --source ${sourceId}; gbrain sources attach ${sourceId}`
: `would: gbrain sources add ${sourceId} --path ${root} --federated; gbrain sync --strategy code --source ${sourceId}; gbrain sources attach ${sourceId}`,
detail: { source_id: sourceId, source_path: root, status: "skipped" },
};
}
@@ -889,10 +933,9 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
// gbrainEnv seeds DATABASE_URL from gbrain's config so this stage works
// inside Next.js / Prisma / Rails projects with their own .env.local
// (codex review #7 — bug fix is wider than #1508 as filed).
const gbrainEnv = buildGbrainEnv({ announce: !args.quiet });
const legacyId = deriveLegacyCodeSourceId(root);
let legacyRemoved = false;
if (legacyId !== sourceId) {
if (!pinnedSourceId && legacyId !== sourceId) {
// #1734: route through the data-loss guards (autopilot + source-safety).
const rm = safeSourcesRemove(legacyId, gbrainEnv);
if (rm.skipped && !args.quiet) {
@@ -908,7 +951,9 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
// pages); fall back to register-new → sync-OK → remove-old. Path-drift
// (user moved the repo, etc.) skips migration with a warning.
const pathOnlyHashLegacyId = derivePathOnlyHashLegacyId(root);
const migration = planHostnameFoldMigration(root, sourceId, pathOnlyHashLegacyId, gbrainEnv);
const migration = pinnedSourceId
? { kind: "none", reason: "no-legacy-source" } as const
: planHostnameFoldMigration(root, sourceId, pathOnlyHashLegacyId, gbrainEnv);
if (migration.kind === "skipped-path-drift" && !args.quiet) {
console.error(
`[sync:code] hostname-fold migration skipped: legacy source ${migration.oldId} `
@@ -919,21 +964,24 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
console.error(`[sync:code] hostname-fold migration: renamed ${migration.oldId}${migration.newId} (pages preserved)`);
}
// Step 1: Ensure source registered (idempotent). Single source of truth in lib —
// no synchronous duplicate here (per /codex review #12).
// Step 1: Ensure generated sources are registered. A confirmed explicit pin
// belongs to the user: its realpath was checked above, so never remove/add it
// merely because the registered spelling differs (e.g. a symlinked checkout).
let registered = false;
try {
const result = await ensureSourceRegistered(sourceId, root, { federated: true, env: gbrainEnv });
registered = result.changed;
} catch (err) {
return {
name: "code",
ran: true,
ok: false,
duration_ms: Date.now() - t0,
summary: `source registration failed: ${(err as Error).message}`,
detail: { source_id: sourceId, source_path: root, status: "failed" },
};
if (!pinnedSourceId) {
try {
const result = await ensureSourceRegistered(sourceId, root, { federated: true, env: gbrainEnv });
registered = result.changed;
} catch (err) {
return {
name: "code",
ran: true,
ok: false,
duration_ms: Date.now() - t0,
summary: `source registration failed: ${(err as Error).message}`,
detail: { source_id: sourceId, source_path: root, status: "failed" },
};
}
}
// Step 2: Always run the page-creating file walk first, then (for --full)
@@ -1336,7 +1384,7 @@ export async function runDream(args: CliArgs): Promise<StageResult> {
if (args.mode === "dry-run") {
const root = repoRoot();
const sourceId = root ? deriveCodeSourceId(root) : null;
const sourceId = root ? readPinnedSourceId(root) ?? deriveCodeSourceId(root) : null;
return {
name: "dream",
ran: false,
@@ -1348,6 +1396,7 @@ export async function runDream(args: CliArgs): Promise<StageResult> {
};
}
const gbrainEnv = buildGbrainEnv({ announce: !args.quiet });
const localStatus = localEngineStatus({ noCache: false });
if (localStatus === "timeout") {
warnProbeTimeout("dream"); // #1964: slow-but-healthy — proceed
@@ -1383,7 +1432,7 @@ export async function runDream(args: CliArgs): Promise<StageResult> {
// code-callers/code-callees for this worktree. Falls back to plain `dream`
// only when we can't derive the source id (not in a git repo).
const root = repoRoot();
const sourceId = root ? deriveCodeSourceId(root) : null;
const sourceId = root ? resolveCodeSourceId(root, gbrainEnv) : null;
const dreamArgs = sourceId ? ["dream", "--source", sourceId] : ["dream"];
// spawnGbrain seeds DATABASE_URL from gbrain's config via buildGbrainEnv.
@@ -1675,7 +1724,8 @@ async function main(): Promise<void> {
let cycle: CycleStatus | null = null;
if (!args.dream && args.mode === "full" && !args.noDream && !args.noCode) {
const root = repoRoot();
cycle = root ? cycleCompleted(deriveCodeSourceId(root), process.env) : "unknown";
const gbrainEnv = buildGbrainEnv({ announce: !args.quiet });
cycle = root ? cycleCompleted(resolveCodeSourceId(root, gbrainEnv), gbrainEnv) : "unknown";
}
if (shouldRunDream(args, cycle)) {
dreamStage = await runDream(args);
+142 -1
View File
@@ -8,7 +8,7 @@
*/
import { describe, it, expect, beforeEach, afterEach } from "bun:test";
import { mkdtempSync, writeFileSync, readFileSync, existsSync, rmSync, mkdirSync, chmodSync } from "fs";
import { mkdtempSync, writeFileSync, readFileSync, existsSync, rmSync, mkdirSync, chmodSync, symlinkSync } from "fs";
import { tmpdir } from "os";
import { join } from "path";
import { spawnSync } from "child_process";
@@ -62,6 +62,14 @@ describe("gstack-gbrain-sync CLI", () => {
expect(source).toContain("localEngineStatus");
});
it("uses GBrain's config environment when resolving dream sources", () => {
const source = readFileSync(SCRIPT, "utf-8");
expect(source).not.toContain("resolveCodeSourceId(root, process.env)");
expect(source).toContain("resolveCodeSourceId(root, gbrainEnv)");
expect(source).toContain("cycleCompleted(resolveCodeSourceId(root, gbrainEnv), gbrainEnv)");
});
it("--dry-run with --code-only reports the code import preview only", () => {
const home = makeTestHome();
const gstackHome = join(home, ".gstack");
@@ -122,6 +130,139 @@ describe("gstack-gbrain-sync CLI", () => {
rmSync(home, { recursive: true, force: true });
});
it("uses a local .gbrain-source in dry-run without spawning gbrain", () => {
const home = makeTestHome();
const gstackHome = join(home, ".gstack");
const bindir = mkdtempSync(join(tmpdir(), "gstack-pinned-source-bin-"));
const repo = mkdtempSync(join(tmpdir(), "gstack-pinned-source-repo-"));
const commandLog = join(home, "gbrain-commands.log");
mkdirSync(gstackHome, { recursive: true });
spawnSync("git", ["init", "--quiet", "-b", "main"], { cwd: repo });
writeFileSync(join(repo, ".gbrain-source"), "client-acme-app\n");
writeFileSync(join(bindir, "gbrain"), `#!/bin/sh
printf '%s\\n' "$*" >> "$GSTACK_TEST_GBRAIN_LOG"
exit 99
`);
chmodSync(join(bindir, "gbrain"), 0o755);
const r = spawnSync("bun", [SCRIPT, "--dry-run", "--code-only", "--quiet"], {
encoding: "utf-8",
timeout: 60000,
cwd: repo,
env: {
...process.env,
HOME: home,
GSTACK_HOME: gstackHome,
GSTACK_TEST_GBRAIN_LOG: commandLog,
PATH: `${bindir}:${process.env.PATH || ""}`,
},
});
expect(r.status).toBe(0);
expect(r.stdout).toContain("gbrain sync --strategy code --source client-acme-app");
expect(r.stdout).not.toContain("gbrain sources add");
expect(r.stdout).not.toContain("--federated");
expect(existsSync(commandLog)).toBe(false);
rmSync(repo, { recursive: true, force: true });
rmSync(bindir, { recursive: true, force: true });
rmSync(home, { recursive: true, force: true });
});
it("keeps a symlink-equivalent pinned source registered as-is", () => {
const home = makeTestHome();
const gstackHome = join(home, ".gstack");
const repo = mkdtempSync(join(tmpdir(), "gstack-pinned-source-repo-"));
const linkDir = mkdtempSync(join(tmpdir(), "gstack-pinned-source-link-"));
const link = join(linkDir, "repo");
const bindir = mkdtempSync(join(tmpdir(), "gstack-pinned-source-bin-"));
const commandLog = join(home, "gbrain-commands.log");
mkdirSync(gstackHome, { recursive: true });
mkdirSync(join(home, ".gbrain"), { recursive: true });
writeFileSync(join(home, ".gbrain", "config.json"), JSON.stringify({ engine: "pglite", database_url: "pglite:///test" }));
spawnSync("git", ["init", "--quiet", "-b", "main"], { cwd: repo });
writeFileSync(join(repo, ".gbrain-source"), "client-acme-app\n");
symlinkSync(repo, link, "dir");
writeFileSync(join(bindir, "gbrain"), `#!/bin/sh
printf '%s\\n' "$*" >> "$GSTACK_TEST_GBRAIN_LOG"
case "$*" in
--version) echo 'gbrain 0.42.0.0' ;;
"sources list --json") echo '{"sources":[{"id":"client-acme-app","local_path":"${link}","page_count":1}]}' ;;
"sync --strategy code --source client-acme-app"|"sources attach client-acme-app") ;;
*) echo "unexpected gbrain command: $*" >&2; exit 1 ;;
esac
`);
chmodSync(join(bindir, "gbrain"), 0o755);
const r = spawnSync("bun", [SCRIPT, "--code-only", "--quiet"], {
encoding: "utf-8",
timeout: 60000,
cwd: link,
env: {
...process.env,
HOME: home,
GSTACK_HOME: gstackHome,
GSTACK_TEST_GBRAIN_LOG: commandLog,
PATH: `${bindir}:${process.env.PATH || ""}`,
},
});
const commands = readFileSync(commandLog, "utf-8");
expect(r.status).toBe(0);
expect(commands).toContain("sync --strategy code --source client-acme-app");
expect(commands).toContain("sources attach client-acme-app");
expect(commands).not.toMatch(/^sources (add|remove) /m);
rmSync(repo, { recursive: true, force: true });
rmSync(linkDir, { recursive: true, force: true });
rmSync(bindir, { recursive: true, force: true });
rmSync(home, { recursive: true, force: true });
});
it("uses a local pin for a dry-run dream without spawning gbrain", () => {
const home = makeTestHome();
const gstackHome = join(home, ".gstack");
const bindir = mkdtempSync(join(tmpdir(), "gstack-pinned-dream-bin-"));
const repo = mkdtempSync(join(tmpdir(), "gstack-pinned-dream-repo-"));
mkdirSync(gstackHome, { recursive: true });
spawnSync("git", ["init", "--quiet", "-b", "main"], { cwd: repo });
writeFileSync(join(repo, ".gbrain-source"), "client-acme-app\n");
writeFileSync(join(bindir, "gbrain"), "#!/bin/sh\nexit 99\n");
chmodSync(join(bindir, "gbrain"), 0o755);
const r = spawnSync("bun", [SCRIPT, "--dry-run", "--dream", "--no-code", "--no-memory", "--no-brain-sync", "--quiet"], {
encoding: "utf-8",
timeout: 60000,
cwd: repo,
env: { ...process.env, HOME: home, GSTACK_HOME: gstackHome, PATH: `${bindir}:${process.env.PATH || ""}` },
});
expect(r.status).toBe(0);
expect(r.stdout).toContain("gbrain dream --source client-acme-app");
rmSync(repo, { recursive: true, force: true });
rmSync(bindir, { recursive: true, force: true });
rmSync(home, { recursive: true, force: true });
});
it("falls back to a derived source when .gbrain-source cannot be read", () => {
const home = makeTestHome();
const gstackHome = join(home, ".gstack");
const repo = mkdtempSync(join(tmpdir(), "gstack-unreadable-pin-repo-"));
mkdirSync(gstackHome, { recursive: true });
spawnSync("git", ["init", "--quiet", "-b", "main"], { cwd: repo });
mkdirSync(join(repo, ".gbrain-source"));
const r = spawnSync("bun", [SCRIPT, "--dry-run", "--code-only", "--quiet"], {
encoding: "utf-8",
timeout: 60000,
cwd: repo,
env: { ...process.env, HOME: home, GSTACK_HOME: gstackHome },
});
expect(r.status).toBe(0);
expect(r.stdout).toMatch(/gbrain sources add gstack-code-/);
rmSync(repo, { recursive: true, force: true });
rmSync(home, { recursive: true, force: true });
});
it("derived source ids are gbrain-valid (≤32 chars, alnum + interior hyphens, no dots) for any remote", () => {
// gbrain enforces source ids to be 1-32 lowercase alnum chars with optional interior
// hyphens. Pre-fix, the slug came from canonicalizeRemote() with only `/` and