mirror of
https://github.com/garrytan/gstack.git
synced 2026-07-18 13:37:23 +02:00
Merge remote-tracking branch 'origin/main' into time-attack/wave-test-infra-isolated-review
This commit is contained in:
t
+14
-11
@@ -134,12 +134,12 @@ lookup_default() {
|
||||
redact_repo_visibility) echo "" ;; # empty → fall through to gh/glab detection
|
||||
redact_prepush_hook) echo "false" ;;
|
||||
# Brain-aware planning (v1.48 / T5+T10+T16). Defaults documented inline:
|
||||
# brain_trust_policy@<hash> — unset on fresh install; setup-gbrain
|
||||
# brain_trust_policy@<endpoint-id> — unset on fresh install; setup-gbrain
|
||||
# writes 'personal' for local engines,
|
||||
# asks the user for remote-ambiguous.
|
||||
# salience_allowlist — empty falls through to
|
||||
# SALIENCE_DEFAULT_ALLOWLIST (D9).
|
||||
# user_slug_at_<hash> — empty triggers resolve-user-slug
|
||||
# user_slug_at_<endpoint-id> — empty triggers resolve-user-slug
|
||||
# fallback chain (D4 A3) on first call.
|
||||
brain_trust_policy*) echo "unset" ;;
|
||||
salience_allowlist) echo "" ;;
|
||||
@@ -260,14 +260,16 @@ resolve_user_slug() {
|
||||
case "${1:-}" in
|
||||
get)
|
||||
KEY="${2:?Usage: gstack-config get <key>}"
|
||||
# Validate key (alphanumeric + underscore + optional @<hash> suffix for
|
||||
# endpoint-namespaced keys introduced by the brain-aware planning layer)
|
||||
if ! printf '%s' "$KEY" | grep -qE '^[a-zA-Z0-9_]+(@[a-f0-9]+)?$'; then
|
||||
echo "Error: key must contain only alphanumeric characters, underscores, and an optional @<hex-hash> suffix" >&2
|
||||
# Validate key (alphanumeric + underscore + optional @<endpoint-id> suffix for
|
||||
# endpoint-namespaced keys introduced by the brain-aware planning layer).
|
||||
# Endpoint ids are sha8/sha16 hex for remote MCP URLs, or the literal
|
||||
# "local" for stdio/PGLite engines (see endpoint_hash).
|
||||
if ! printf '%s' "$KEY" | grep -qE '^[a-zA-Z0-9_]+(@[a-zA-Z0-9]+)?$'; then
|
||||
echo "Error: key must contain only alphanumeric characters, underscores, and an optional @<endpoint-id> suffix" >&2
|
||||
exit 1
|
||||
fi
|
||||
# Use literal match for keys containing @ (sha hashes), regex otherwise
|
||||
VALUE=$(grep -F "${KEY}:" "$CONFIG_FILE" 2>/dev/null | grep -E "^${KEY%@*}(@[a-f0-9]+)?:" | grep -F "${KEY}:" | tail -1 | awk '{print $2}' | tr -d '[:space:]' || true)
|
||||
# Use literal match for keys containing @ (endpoint ids), regex otherwise
|
||||
VALUE=$(grep -F "${KEY}:" "$CONFIG_FILE" 2>/dev/null | grep -E "^${KEY%@*}(@[a-zA-Z0-9]+)?:" | grep -F "${KEY}:" | tail -1 | awk '{print $2}' | tr -d '[:space:]' || true)
|
||||
if [ -z "$VALUE" ]; then
|
||||
VALUE=$(lookup_default "$KEY")
|
||||
fi
|
||||
@@ -276,9 +278,10 @@ case "${1:-}" in
|
||||
set)
|
||||
KEY="${2:?Usage: gstack-config set <key> <value>}"
|
||||
VALUE="${3:?Usage: gstack-config set <key> <value>}"
|
||||
# Validate key (alphanumeric + underscore + optional @<hash> suffix)
|
||||
if ! printf '%s' "$KEY" | grep -qE '^[a-zA-Z0-9_]+(@[a-f0-9]+)?$'; then
|
||||
echo "Error: key must contain only alphanumeric characters, underscores, and an optional @<hex-hash> suffix" >&2
|
||||
# Validate key (alphanumeric + underscore + optional @<endpoint-id> suffix).
|
||||
# Accepts hex hashes and the literal "local" from endpoint_hash.
|
||||
if ! printf '%s' "$KEY" | grep -qE '^[a-zA-Z0-9_]+(@[a-zA-Z0-9]+)?$'; then
|
||||
echo "Error: key must contain only alphanumeric characters, underscores, and an optional @<endpoint-id> suffix" >&2
|
||||
exit 1
|
||||
fi
|
||||
# Validate brain_trust_policy value domain (D4 / D11)
|
||||
|
||||
@@ -18,7 +18,7 @@
|
||||
* "gstack_brain_sync_mode": "off"|"artifacts-only"|"full",
|
||||
* "gstack_brain_git": true|false,
|
||||
* "gstack_artifacts_remote": "https://..." | "",
|
||||
* "gbrain_local_status": "ok"|"no-cli"|"missing-config"|"broken-config"|"broken-db"|"timeout",
|
||||
* "gbrain_local_status": "ok"|"no-cli"|"missing-config"|"broken-config"|"broken-db"|"engine-locked"|"timeout",
|
||||
* "gbrain_pooler_mode": "transaction"|"session"|null
|
||||
* }
|
||||
*
|
||||
|
||||
@@ -717,6 +717,7 @@ function dreamMarkerPid(): number | null {
|
||||
* missing-config → "no local engine; run /setup-gbrain to add local PGLite"
|
||||
* broken-config → "config file at ~/.gbrain/config.json is malformed; see /setup-gbrain Step 1.5"
|
||||
* broken-db → "config points at unreachable DB; see /setup-gbrain Step 1.5"
|
||||
* engine-locked → PGLite is busy; stop its holder or sync outside the live session
|
||||
* timeout → kept for Record totality; stages PROCEED on timeout (#1964)
|
||||
* via the gate's warnProbeTimeout path, never this skip.
|
||||
*/
|
||||
@@ -733,6 +734,8 @@ function skipStageForLocalStatus(
|
||||
"config at ~/.gbrain/config.json is malformed; see /setup-gbrain Step 1.5",
|
||||
"broken-db":
|
||||
"config points at unreachable DB; see /setup-gbrain Step 1.5",
|
||||
"engine-locked":
|
||||
"PGLite is busy (often held by gbrain serve); stop the holding process or run /sync-gbrain outside the live Claude session, then retry",
|
||||
"timeout":
|
||||
"engine probe timed out; raise GSTACK_GBRAIN_PROBE_TIMEOUT_MS if your pooler is slow",
|
||||
};
|
||||
|
||||
@@ -12,13 +12,20 @@
|
||||
set -euo pipefail
|
||||
|
||||
SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
|
||||
# Compute SLUG directly (avoid eval of gstack-slug — branch names can contain shell metacharacters)
|
||||
# Keep the no-remote early exit before consulting gstack-slug. gstack-slug falls
|
||||
# back to the directory basename when origin is missing, which would incorrectly
|
||||
# make local-only repos eligible for classification.
|
||||
REMOTE_URL=$(git remote get-url origin 2>/dev/null || true)
|
||||
if [ -z "$REMOTE_URL" ]; then
|
||||
echo "REPO_MODE=unknown"
|
||||
exit 0
|
||||
fi
|
||||
SLUG=$(echo "$REMOTE_URL" | sed 's|.*[:/]\([^/]*/[^/]*\)\.git$|\1|;s|.*[:/]\([^/]*/[^/]*\)$|\1|' | tr '/' '-')
|
||||
# Consume the canonical cached slug from gstack-slug (sed, never eval — branch
|
||||
# names can contain shell metacharacters). Invoke from the git root so the
|
||||
# PWD-keyed slug cache matches other gstack project-state writers.
|
||||
REPO_ROOT=$(git rev-parse --show-toplevel 2>/dev/null || true)
|
||||
[ -z "${REPO_ROOT:-}" ] && { echo "REPO_MODE=unknown"; exit 0; }
|
||||
SLUG=$(cd "$REPO_ROOT" && "$SCRIPT_DIR/gstack-slug" | sed -n 's/^SLUG=//p' | head -1)
|
||||
[ -z "${SLUG:-}" ] && { echo "REPO_MODE=unknown"; exit 0; }
|
||||
|
||||
# Validate: only allow known values (prevent shell injection via source <(...))
|
||||
|
||||
@@ -338,19 +338,28 @@ const BLOCKLIST_DOMAINS = [
|
||||
export function urlBlocklistFilter(content: string, url: string, _command: string): ContentFilterResult {
|
||||
const warnings: string[] = [];
|
||||
|
||||
// Matching is case-insensitive. Hostnames and URL schemes are
|
||||
// case-insensitive (RFC 3986), so an uppercased sink like
|
||||
// https://WEBHOOK.SITE/x must not slip past the blocklist. Normalize the
|
||||
// haystack to lowercase; BLOCKLIST_DOMAINS entries are lowercase by
|
||||
// construction, so a direct compare then holds.
|
||||
|
||||
// Check page URL
|
||||
const normalizedUrl = url.toLowerCase();
|
||||
for (const domain of BLOCKLIST_DOMAINS) {
|
||||
if (url.includes(domain)) {
|
||||
if (normalizedUrl.includes(domain)) {
|
||||
warnings.push(`Page URL matches blocklisted domain: ${domain}`);
|
||||
}
|
||||
}
|
||||
|
||||
// Check for blocklisted URLs in content (links, form actions)
|
||||
const urlPattern = /https?:\/\/[^\s"'<>]+/g;
|
||||
// Check for blocklisted URLs in content (links, form actions). The `i` flag
|
||||
// keeps an uppercased scheme (HTTPS://) from evading URL extraction.
|
||||
const urlPattern = /https?:\/\/[^\s"'<>]+/gi;
|
||||
const contentUrls = content.match(urlPattern) || [];
|
||||
for (const contentUrl of contentUrls) {
|
||||
const normalizedContentUrl = contentUrl.toLowerCase();
|
||||
for (const domain of BLOCKLIST_DOMAINS) {
|
||||
if (contentUrl.includes(domain)) {
|
||||
if (normalizedContentUrl.includes(domain)) {
|
||||
warnings.push(`Content contains blocklisted URL: ${contentUrl.slice(0, 100)}`);
|
||||
break;
|
||||
}
|
||||
|
||||
@@ -151,6 +151,44 @@ describe('Content filter hooks', () => {
|
||||
expect(result.warnings.length).toBe(0);
|
||||
});
|
||||
|
||||
// Regression: issue #2190 — case-sensitive matching let an uppercased
|
||||
// exfiltration domain bypass the blocklist.
|
||||
test('URL blocklist is case-insensitive on the page URL', () => {
|
||||
const result = urlBlocklistFilter('', 'https://WEBHOOK.SITE/steal', 'text');
|
||||
expect(result.safe).toBe(false);
|
||||
expect(result.warnings.some(w => w.includes('webhook.site'))).toBe(true);
|
||||
});
|
||||
|
||||
test('URL blocklist is case-insensitive on content URLs', () => {
|
||||
const result = urlBlocklistFilter(
|
||||
'<a href="https://WEBHOOK.SITE/a1b2c3-steal">click</a>',
|
||||
'https://docs.example.com/article',
|
||||
'links',
|
||||
);
|
||||
expect(result.safe).toBe(false);
|
||||
expect(result.warnings.some(w => w.includes('WEBHOOK.SITE'))).toBe(true);
|
||||
});
|
||||
|
||||
test('URL blocklist catches an uppercased scheme in content', () => {
|
||||
const result = urlBlocklistFilter(
|
||||
'Exfil via HTTPS://Requestbin.com/r/abc please',
|
||||
'https://example.com',
|
||||
'text',
|
||||
);
|
||||
expect(result.safe).toBe(false);
|
||||
expect(result.warnings.some(w => w.toLowerCase().includes('requestbin.com'))).toBe(true);
|
||||
});
|
||||
|
||||
test('URL blocklist does not over-block legitimate uppercase domains', () => {
|
||||
const result = urlBlocklistFilter(
|
||||
'<a href="https://GitHub.com/acme/project">source</a>',
|
||||
'https://DOCS.EXAMPLE.COM/help',
|
||||
'links',
|
||||
);
|
||||
expect(result.safe).toBe(true);
|
||||
expect(result.warnings.length).toBe(0);
|
||||
});
|
||||
|
||||
test('custom filter can be registered and runs', () => {
|
||||
registerContentFilter((content, url, cmd) => {
|
||||
if (content.includes('SECRET')) {
|
||||
|
||||
@@ -108,6 +108,13 @@ describe('gstack-config', () => {
|
||||
expect(existsSync(join(nestedDir, 'config.yaml'))).toBe(true);
|
||||
});
|
||||
|
||||
test('brain trust policy accepts local endpoint suffix', () => {
|
||||
const { exitCode, stderr } = run(['set', 'brain_trust_policy@local', 'personal']);
|
||||
expect(exitCode).toBe(0);
|
||||
expect(stderr).toBe('');
|
||||
expect(run(['get', 'brain_trust_policy@local']).stdout).toBe('personal');
|
||||
});
|
||||
|
||||
// ─── list ─────────────────────────────────────────────────
|
||||
test('list shows all keys', () => {
|
||||
writeFileSync(join(stateDir, 'config.yaml'), 'auto_upgrade: true\nupdate_check: false\n');
|
||||
@@ -139,6 +146,12 @@ describe('gstack-config', () => {
|
||||
expect(stderr).toContain('alphanumeric');
|
||||
});
|
||||
|
||||
test('set rejects endpoint suffix with punctuation', () => {
|
||||
const { exitCode, stderr } = run(['set', 'brain_trust_policy@local-dev', 'personal']);
|
||||
expect(exitCode).toBe(1);
|
||||
expect(stderr).toContain('endpoint-id');
|
||||
});
|
||||
|
||||
test('set preserves value with sed special chars', () => {
|
||||
run(['set', 'test_special', 'a/b&c\\d']);
|
||||
const { stdout } = run(['get', 'test_special']);
|
||||
|
||||
@@ -22,6 +22,9 @@
|
||||
},
|
||||
},
|
||||
},
|
||||
"overrides": {
|
||||
"basic-ftp": "5.3.1",
|
||||
},
|
||||
"packages": {
|
||||
"@anthropic-ai/claude-agent-sdk": ["@anthropic-ai/claude-agent-sdk@0.2.117", "", { "dependencies": { "@anthropic-ai/sdk": "^0.81.0", "@modelcontextprotocol/sdk": "^1.29.0" }, "optionalDependencies": { "@anthropic-ai/claude-agent-sdk-darwin-arm64": "0.2.117", "@anthropic-ai/claude-agent-sdk-darwin-x64": "0.2.117", "@anthropic-ai/claude-agent-sdk-linux-arm64": "0.2.117", "@anthropic-ai/claude-agent-sdk-linux-arm64-musl": "0.2.117", "@anthropic-ai/claude-agent-sdk-linux-x64": "0.2.117", "@anthropic-ai/claude-agent-sdk-linux-x64-musl": "0.2.117", "@anthropic-ai/claude-agent-sdk-win32-arm64": "0.2.117", "@anthropic-ai/claude-agent-sdk-win32-x64": "0.2.117" }, "peerDependencies": { "zod": "^4.0.0" } }, "sha512-pVBss1Vu0w87nKCBhWtjMggSgCh6GVUtdRmuE58ZvXv0E2q0JcnUCQHehmn92BAW0+VCwPY8q/k7uKWkgwz/gA=="],
|
||||
|
||||
@@ -201,7 +204,7 @@
|
||||
|
||||
"bare-url": ["bare-url@2.4.0", "", { "dependencies": { "bare-path": "^3.0.0" } }, "sha512-NSTU5WN+fy/L0DDenfE8SXQna4voXuW0FHM7wH8i3/q9khUSchfPbPezO4zSFMnDGIf9YE+mt/RWhZgNRKRIXA=="],
|
||||
|
||||
"basic-ftp": ["basic-ftp@5.2.0", "", {}, "sha512-VoMINM2rqJwJgfdHq6RiUudKt2BV+FY5ZFezP/ypmwayk68+NzzAQy4XXLlqsGD4MCzq3DrmNFD/uUmBJuGoXw=="],
|
||||
"basic-ftp": ["basic-ftp@5.3.1", "", {}, "sha512-bopVNp6ugyA150DDuZfPFdt1KZ5a94ZDiwX4hMgZDzF+GttD80lEy8kj98kbyhLXnPvhtIo93mdnLIjpCAeeOw=="],
|
||||
|
||||
"body-parser": ["body-parser@2.2.2", "", { "dependencies": { "bytes": "^3.1.2", "content-type": "^1.0.5", "debug": "^4.4.3", "http-errors": "^2.0.0", "iconv-lite": "^0.7.0", "on-finished": "^2.4.1", "qs": "^6.14.1", "raw-body": "^3.0.1", "type-is": "^2.0.1" } }, "sha512-oP5VkATKlNwcgvxi0vM0p/D3n2C3EReYVX+DNYs5TjZFn/oQt2j+4sVJtSMr18pdRr8wjTcBl6LoV+FUwzPmNA=="],
|
||||
|
||||
|
||||
@@ -19,6 +19,8 @@
|
||||
* Broken-config → config exists but `gbrain sources list` fails with config parse error
|
||||
* (or any non-recognized error — defensive default per codex #8).
|
||||
* Broken-db → config exists, DB unreachable per stderr classification.
|
||||
* Engine-locked → PGLite probe hit gbrain's own connect timeout, usually
|
||||
* because another `gbrain serve` process owns the embedded DB.
|
||||
* Timeout → probe exceeded GSTACK_GBRAIN_PROBE_TIMEOUT_MS (default 15s) with no
|
||||
* recognized error — engine is likely healthy but slow (e.g. a cold
|
||||
* pooler connection, #1964). Consumers treat this as usable.
|
||||
@@ -47,6 +49,7 @@ export type LocalEngineStatus =
|
||||
| "missing-config"
|
||||
| "broken-config"
|
||||
| "broken-db"
|
||||
| "engine-locked"
|
||||
| "timeout";
|
||||
|
||||
export interface ClassifyOptions {
|
||||
@@ -118,6 +121,15 @@ function gbrainConfigPath(env?: NodeJS.ProcessEnv): string {
|
||||
return join(gbrainHome, "config.json");
|
||||
}
|
||||
|
||||
function configuredEngine(env?: NodeJS.ProcessEnv): "pglite" | "postgres" | null {
|
||||
try {
|
||||
const parsed = JSON.parse(readFileSync(gbrainConfigPath(env), "utf-8")) as { engine?: string };
|
||||
return parsed.engine === "pglite" || parsed.engine === "postgres" ? parsed.engine : null;
|
||||
} catch {
|
||||
return null;
|
||||
}
|
||||
}
|
||||
|
||||
function hashPath(p: string): string {
|
||||
return createHash("sha256").update(p).digest("hex").slice(0, 16);
|
||||
}
|
||||
@@ -281,6 +293,7 @@ function freshClassify(env?: NodeJS.ProcessEnv): LocalEngineStatus {
|
||||
stderr?: Buffer | string;
|
||||
killed?: boolean;
|
||||
signal?: NodeJS.Signals | null;
|
||||
status?: number | null;
|
||||
};
|
||||
const stderr = (e.stderr ? e.stderr.toString() : "") || "";
|
||||
|
||||
@@ -292,6 +305,14 @@ function freshClassify(env?: NodeJS.ProcessEnv): LocalEngineStatus {
|
||||
if (stderr.includes("Cannot connect to database")) return "broken-db";
|
||||
if (stderr.includes("config.json")) return "broken-config";
|
||||
|
||||
// PGLite is single-process. A long-lived `gbrain serve` can own the
|
||||
// embedded database, causing the CLI to finish with its own exit 124 and
|
||||
// "connect timed out" message. This is neither our watchdog timeout nor
|
||||
// evidence that the valid config is malformed (#2194).
|
||||
if (stderr.includes("connect timed out") || e.status === 124) {
|
||||
return configuredEngine(env) === "pglite" ? "engine-locked" : "broken-db";
|
||||
}
|
||||
|
||||
// Probe killed by the timeout with no recognized error: the engine is
|
||||
// most likely healthy but slow (cold pooler connections measured at
|
||||
// 6.9-10.7s in #1964). Don't tell the user their config is malformed.
|
||||
|
||||
@@ -76,5 +76,8 @@
|
||||
"@anthropic-ai/sdk": "^0.78.0",
|
||||
"xterm": "5",
|
||||
"xterm-addon-fit": "^0.8.0"
|
||||
},
|
||||
"overrides": {
|
||||
"basic-ftp": "5.3.1"
|
||||
}
|
||||
}
|
||||
|
||||
@@ -820,7 +820,7 @@ Capture the JSON output. It contains: `gbrain_on_path`, `gbrain_version`,
|
||||
`gbrain_config_exists`, `gbrain_engine`, `gbrain_doctor_ok`, `gbrain_mcp_mode`,
|
||||
`gstack_brain_sync_mode`, `gstack_brain_git`, `gstack_artifacts_remote`, and
|
||||
the v1.34.0.0+ `gbrain_local_status` field (one of: `ok`, `no-cli`,
|
||||
`missing-config`, `broken-config`, `broken-db`, `timeout`). Treat `timeout`
|
||||
`missing-config`, `broken-config`, `broken-db`, `engine-locked`, `timeout`). Treat `timeout`
|
||||
like `ok` (slow-but-healthy engine, #1964) — it never triggers Step 1.5
|
||||
remediation.
|
||||
|
||||
|
||||
@@ -66,7 +66,7 @@ Capture the JSON output. It contains: `gbrain_on_path`, `gbrain_version`,
|
||||
`gbrain_config_exists`, `gbrain_engine`, `gbrain_doctor_ok`, `gbrain_mcp_mode`,
|
||||
`gstack_brain_sync_mode`, `gstack_brain_git`, `gstack_artifacts_remote`, and
|
||||
the v1.34.0.0+ `gbrain_local_status` field (one of: `ok`, `no-cli`,
|
||||
`missing-config`, `broken-config`, `broken-db`, `timeout`). Treat `timeout`
|
||||
`missing-config`, `broken-config`, `broken-db`, `engine-locked`, `timeout`). Treat `timeout`
|
||||
like `ok` (slow-but-healthy engine, #1964) — it never triggers Step 1.5
|
||||
remediation.
|
||||
|
||||
|
||||
@@ -893,6 +893,10 @@ BEFORE invoking the orchestrator:
|
||||
slow (cold pooler connection, #1964). Tell the user in one line: "Engine
|
||||
probe timed out (>15s) — proceeding; raise `GSTACK_GBRAIN_PROBE_TIMEOUT_MS`
|
||||
if your pooler is slow." Do NOT treat this as a broken config.
|
||||
- **`engine-locked`**: STOP. "The local PGLite database is busy, usually
|
||||
because `gbrain serve` from a live Claude session owns it. Stop that process
|
||||
or run `/sync-gbrain` outside the live session, then retry. This identifies
|
||||
the conflict but does not remove PGLite's single-process limit."
|
||||
- **`no-cli`**: STOP. "Local gbrain CLI not installed. Run `/setup-gbrain`
|
||||
first."
|
||||
- **`missing-config`** AND `gbrain_mcp_mode == "remote-http"`: tell the user
|
||||
|
||||
@@ -139,6 +139,10 @@ BEFORE invoking the orchestrator:
|
||||
slow (cold pooler connection, #1964). Tell the user in one line: "Engine
|
||||
probe timed out (>15s) — proceeding; raise `GSTACK_GBRAIN_PROBE_TIMEOUT_MS`
|
||||
if your pooler is slow." Do NOT treat this as a broken config.
|
||||
- **`engine-locked`**: STOP. "The local PGLite database is busy, usually
|
||||
because `gbrain serve` from a live Claude session owns it. Stop that process
|
||||
or run `/sync-gbrain` outside the live session, then retry. This identifies
|
||||
the conflict but does not remove PGLite's single-process limit."
|
||||
- **`no-cli`**: STOP. "Local gbrain CLI not installed. Run `/setup-gbrain`
|
||||
first."
|
||||
- **`missing-config`** AND `gbrain_mcp_mode == "remote-http"`: tell the user
|
||||
|
||||
@@ -0,0 +1,58 @@
|
||||
import { describe, test, expect } from 'bun:test';
|
||||
import { readFileSync } from 'fs';
|
||||
import { join } from 'path';
|
||||
|
||||
const ROOT = join(import.meta.dir, '..');
|
||||
|
||||
// Security regression guard for the basic-ftp transitive dependency.
|
||||
//
|
||||
// basic-ftp reaches the tree only transitively:
|
||||
// puppeteer-core > @puppeteer/browsers > proxy-agent > pac-proxy-agent > get-uri > basic-ftp
|
||||
//
|
||||
// Versions <= 5.3.0 carry four HIGH advisories, all fixed in 5.3.1:
|
||||
// - GHSA-chqc-8p9q-pq6q CVE-2026-39983 FTP command injection via CRLF (fixed 5.2.1)
|
||||
// - GHSA-6v7q-wjvx-w8wg incomplete CRLF protection, USER/PASS + MKD bypass (fixed 5.2.2)
|
||||
// - GHSA-rpmf-866q-6p89 DoS via unbounded multiline control-response buffering
|
||||
// - GHSA-rp42-5vxx-qpwr DoS via unbounded memory in Client.list()
|
||||
//
|
||||
// The fix is a bun `overrides` pin (not a phantom direct dependency): overriding
|
||||
// forces EVERY basic-ftp in the tree to the safe version, including get-uri's
|
||||
// nested copy. A direct-dependency bump leaves that nested copy behind — which is
|
||||
// exactly the failure mode this test is here to catch. See the CVE-2026-39983
|
||||
// upgrade fix for the full rationale.
|
||||
const MIN_SAFE = '5.3.1';
|
||||
|
||||
function cmpSemver(a: string, b: string): number {
|
||||
const pa = a.split('.').map(Number);
|
||||
const pb = b.split('.').map(Number);
|
||||
for (let i = 0; i < 3; i++) {
|
||||
if ((pa[i] ?? 0) !== (pb[i] ?? 0)) return (pa[i] ?? 0) - (pb[i] ?? 0);
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
|
||||
describe('basic-ftp security pin (CVE-2026-39983 and siblings)', () => {
|
||||
test('package.json overrides basic-ftp to a safe version', () => {
|
||||
const pkg = JSON.parse(readFileSync(join(ROOT, 'package.json'), 'utf-8'));
|
||||
const pin = pkg.overrides?.['basic-ftp'];
|
||||
expect(pin, 'package.json overrides.basic-ftp must exist').toBeDefined();
|
||||
const pinned = String(pin).replace(/^[\^~]/, '');
|
||||
expect(
|
||||
cmpSemver(pinned, MIN_SAFE) >= 0,
|
||||
`overrides.basic-ftp is "${pin}" but must be >= ${MIN_SAFE}`,
|
||||
).toBe(true);
|
||||
});
|
||||
|
||||
test('no basic-ftp entry in bun.lock resolves below the safe version', () => {
|
||||
const lock = readFileSync(join(ROOT, 'bun.lock'), 'utf-8');
|
||||
// Match every resolved basic-ftp specifier, including nested paths like
|
||||
// "get-uri/basic-ftp" that a direct-dependency bump would leave vulnerable.
|
||||
const versions = [...lock.matchAll(/basic-ftp@(\d+\.\d+\.\d+)/g)].map(m => m[1]);
|
||||
expect(versions.length, 'expected at least one basic-ftp entry in bun.lock').toBeGreaterThan(0);
|
||||
const vulnerable = versions.filter(v => cmpSemver(v, MIN_SAFE) < 0);
|
||||
expect(
|
||||
vulnerable,
|
||||
`bun.lock still resolves vulnerable basic-ftp version(s): ${vulnerable.join(', ')}`,
|
||||
).toEqual([]);
|
||||
});
|
||||
});
|
||||
@@ -6,13 +6,14 @@
|
||||
* on PATH that emits canned exit codes + stderr matching the patterns the
|
||||
* classifier looks for.
|
||||
*
|
||||
* Six status cases:
|
||||
* Seven status cases:
|
||||
* 1. no-cli — gbrain absent from PATH
|
||||
* 2. missing-config — gbrain present, config.json absent (honors GBRAIN_HOME)
|
||||
* 3. broken-config — gbrain present, config exists, stderr contains "config.json"
|
||||
* 4. broken-db — gbrain present, config exists, stderr contains "Cannot connect to database"
|
||||
* 5. timeout — probe exceeds GSTACK_GBRAIN_PROBE_TIMEOUT_MS with no recognized error (#1964)
|
||||
* 6. ok — gbrain present, config exists, sources list returns valid JSON
|
||||
* 6. engine-locked — PGLite CLI exits 124 because another process owns the DB (#2194)
|
||||
* 7. ok — gbrain present, config exists, sources list returns valid JSON
|
||||
*
|
||||
* Plus cache behavior: hit, TTL expiry, invariant invalidation (HOME change,
|
||||
* probe-timeout change), --no-cache bypass. Timeout tests keep runtime sane by
|
||||
@@ -61,7 +62,7 @@ interface FakeEnv {
|
||||
*/
|
||||
function makeEnv(opts: {
|
||||
withGbrain?: boolean;
|
||||
gbrainBehavior?: "ok" | "broken-db" | "broken-config" | "throws" | "slow";
|
||||
gbrainBehavior?: "ok" | "broken-db" | "broken-config" | "engine-locked" | "throws" | "slow";
|
||||
withConfig?: boolean;
|
||||
}): FakeEnv {
|
||||
const tmp = mkdtempSync(join(tmpdir(), "gbrain-local-status-test-"));
|
||||
@@ -102,7 +103,7 @@ function makeEnv(opts: {
|
||||
}
|
||||
|
||||
function makeFakeGbrainScript(
|
||||
behavior: "ok" | "broken-db" | "broken-config" | "throws" | "slow",
|
||||
behavior: "ok" | "broken-db" | "broken-config" | "engine-locked" | "throws" | "slow",
|
||||
): string {
|
||||
// "slow": healthy engine on a cold pooler connection (#1964) — sleeps past
|
||||
// the (test-lowered) probe timeout, then would answer fine.
|
||||
@@ -125,10 +126,12 @@ exit 0
|
||||
? 'echo "Cannot connect to database: . Fix: Check your connection URL in ~/.gbrain/config.json" >&2'
|
||||
: behavior === "broken-config"
|
||||
? 'echo "Error: malformed config.json at ~/.gbrain/config.json" >&2'
|
||||
: behavior === "engine-locked"
|
||||
? 'echo "gbrain sources: connect timed out (default 10000ms; pass --timeout=Ns to override)." >&2'
|
||||
: behavior === "throws"
|
||||
? 'echo "unexpected gbrain failure" >&2'
|
||||
: "";
|
||||
const exitCode = behavior === "ok" ? 0 : 1;
|
||||
const exitCode = behavior === "ok" ? 0 : behavior === "engine-locked" ? 124 : 1;
|
||||
return `#!/bin/sh
|
||||
if [ "$1" = "--version" ]; then
|
||||
echo "gbrain 0.33.1.0"
|
||||
@@ -225,6 +228,19 @@ describe("lib/gbrain-local-status — status classification", () => {
|
||||
expect(localEngineStatus({ noCache: true })).toBe("broken-config");
|
||||
});
|
||||
|
||||
it("returns 'engine-locked' when PGLite exits 124 with its own connect timeout (#2194)", () => {
|
||||
env = makeEnv({ withGbrain: true, gbrainBehavior: "engine-locked", withConfig: true });
|
||||
restoreEnv = applyEnv(env);
|
||||
expect(localEngineStatus({ noCache: true })).toBe("engine-locked");
|
||||
});
|
||||
|
||||
it("classifies a non-PGLite connect timeout as unreachable DB, not malformed config", () => {
|
||||
env = makeEnv({ withGbrain: true, gbrainBehavior: "engine-locked", withConfig: true });
|
||||
restoreEnv = applyEnv(env);
|
||||
writeFileSync(env.configPath, JSON.stringify({ engine: "postgres", database_url: "postgres://fake" }));
|
||||
expect(localEngineStatus({ noCache: true })).toBe("broken-db");
|
||||
});
|
||||
|
||||
it("returns 'ok' when sources list succeeds", () => {
|
||||
env = makeEnv({ withGbrain: true, gbrainBehavior: "ok", withConfig: true });
|
||||
restoreEnv = applyEnv(env);
|
||||
|
||||
@@ -42,7 +42,7 @@ interface FakeEnv {
|
||||
*/
|
||||
function makeEnv(opts: {
|
||||
withGbrain: boolean;
|
||||
gbrainBehavior?: "ok" | "broken-db" | "broken-config" | "slow";
|
||||
gbrainBehavior?: "ok" | "broken-db" | "broken-config" | "engine-locked" | "slow";
|
||||
withConfig: boolean;
|
||||
}): FakeEnv {
|
||||
const tmp = mkdtempSync(join(tmpdir(), "gbrain-sync-skip-"));
|
||||
@@ -75,6 +75,9 @@ function makeEnv(opts: {
|
||||
: behavior === "ok"
|
||||
? ` echo '{"sources":[]}'
|
||||
exit 0`
|
||||
: behavior === "engine-locked"
|
||||
? ` echo "gbrain sources: connect timed out (default 10000ms; pass --timeout=Ns to override)." >&2
|
||||
exit 124`
|
||||
: ` ${
|
||||
behavior === "broken-db"
|
||||
? 'echo "Cannot connect to database: . Fix: Check your connection URL in ~/.gbrain/config.json" >&2'
|
||||
@@ -207,6 +210,20 @@ describe("gstack-gbrain-sync — split-engine SKIP (plan D12)", () => {
|
||||
}
|
||||
});
|
||||
|
||||
it("SKIPs with actionable guidance when PGLite is held by gbrain serve (#2194)", () => {
|
||||
const env = makeEnv({ withGbrain: true, gbrainBehavior: "engine-locked", withConfig: true });
|
||||
try {
|
||||
const r = runOrchestrator(env, ["--code-only"]);
|
||||
const out = r.stdout + r.stderr;
|
||||
expect(out).toContain("local engine engine-locked");
|
||||
expect(out).toContain("gbrain serve");
|
||||
expect(out).toContain("outside the live Claude session");
|
||||
expect(out).not.toContain("config.json is malformed");
|
||||
} finally {
|
||||
env.cleanup();
|
||||
}
|
||||
});
|
||||
|
||||
it("SKIPs code stage when gbrain CLI is missing (no-cli)", () => {
|
||||
const env = makeEnv({ withGbrain: false, withConfig: false });
|
||||
try {
|
||||
|
||||
@@ -0,0 +1,176 @@
|
||||
/**
|
||||
* Project-slug consistency for gstack-repo-mode.
|
||||
*
|
||||
* These are end-to-end reproductions of #2212. Each case is a real local Git
|
||||
* workspace which predates its public GitHub origin: gstack-slug caches the
|
||||
* local directory name, then the origin is added. repo-mode must write its
|
||||
* cache next to that canonical project state, not beside a newly derived
|
||||
* owner-repo slug.
|
||||
*/
|
||||
|
||||
import { describe, expect, test } from 'bun:test';
|
||||
import { spawnSync } from 'child_process';
|
||||
import fs from 'fs';
|
||||
import os from 'os';
|
||||
import path from 'path';
|
||||
|
||||
const ROOT = path.resolve(import.meta.dir, '..');
|
||||
const REPO_MODE_BIN = path.join(ROOT, 'bin', 'gstack-repo-mode');
|
||||
const SLUG_BIN = path.join(ROOT, 'bin', 'gstack-slug');
|
||||
|
||||
type CommandResult = { stdout: string; stderr: string; status: number };
|
||||
|
||||
function run(command: string, args: string[], cwd: string, home: string): CommandResult {
|
||||
const result = spawnSync(command, args, {
|
||||
cwd,
|
||||
encoding: 'utf8',
|
||||
env: { ...process.env, HOME: home, GSTACK_HOME: path.join(home, '.gstack') },
|
||||
});
|
||||
return {
|
||||
stdout: result.stdout ?? '',
|
||||
stderr: result.stderr ?? '',
|
||||
status: result.status ?? -1,
|
||||
};
|
||||
}
|
||||
|
||||
function git(args: string[], cwd: string, home: string) {
|
||||
const result = run('git', args, cwd, home);
|
||||
expect(result.status, `${result.stderr}\n${result.stdout}`).toBe(0);
|
||||
}
|
||||
|
||||
function slugFrom(output: string): string {
|
||||
const line = output.split('\n').find((entry) => entry.startsWith('SLUG='));
|
||||
expect(line).toBeDefined();
|
||||
return line!.slice('SLUG='.length);
|
||||
}
|
||||
|
||||
const publicWorkspaces = [
|
||||
{ name: 'react-local-history', origin: 'https://github.com/facebook/react.git' },
|
||||
{ name: 'next-local-history', origin: 'https://github.com/vercel/next.js.git' },
|
||||
{ name: 'kubernetes-local-history', origin: 'https://github.com/kubernetes/kubernetes.git' },
|
||||
];
|
||||
|
||||
describe('gstack-repo-mode cached slug consistency (#2212)', () => {
|
||||
for (const workspace of publicWorkspaces) {
|
||||
test(`${workspace.origin} keeps its pre-origin project cache`, () => {
|
||||
const home = fs.mkdtempSync(path.join(os.tmpdir(), 'grepo-home-'));
|
||||
const root = fs.mkdtempSync(path.join(os.tmpdir(), 'grepo-workspaces-'));
|
||||
const project = path.join(root, workspace.name);
|
||||
|
||||
try {
|
||||
fs.mkdirSync(project);
|
||||
git(['init', '-b', 'main'], project, home);
|
||||
git(['config', 'user.name', 'gstack regression test'], project, home);
|
||||
git(['config', 'user.email', 'test@gstack.dev'], project, home);
|
||||
|
||||
// This workspace existed and used gstack before it adopted GitHub.
|
||||
const initialSlug = slugFrom(run(SLUG_BIN, [], project, home).stdout);
|
||||
expect(initialSlug).toBe(workspace.name);
|
||||
|
||||
// Five commits make repo-mode perform its real classification path.
|
||||
for (let commit = 1; commit <= 5; commit += 1) {
|
||||
fs.writeFileSync(path.join(project, `commit-${commit}.txt`), String(commit));
|
||||
git(['add', '.'], project, home);
|
||||
git(['commit', '-m', `commit ${commit}`], project, home);
|
||||
}
|
||||
git(['remote', 'add', 'origin', workspace.origin], project, home);
|
||||
|
||||
const mode = run(REPO_MODE_BIN, [], project, home);
|
||||
expect(mode.status, mode.stderr).toBe(0);
|
||||
expect(mode.stdout.trim()).toBe('REPO_MODE=solo');
|
||||
|
||||
// Canonical slug stays the pre-origin basename; remote-derived twin must not appear.
|
||||
const projectsDir = path.join(home, '.gstack', 'projects');
|
||||
const canonicalCache = path.join(projectsDir, initialSlug, 'repo-mode.json');
|
||||
expect(fs.existsSync(canonicalCache)).toBe(true);
|
||||
expect(fs.readdirSync(projectsDir)).toEqual([initialSlug]);
|
||||
|
||||
// Slug remains sanitized even when the remote URL would otherwise diverge.
|
||||
expect(initialSlug).toMatch(/^[a-zA-Z0-9._-]+$/);
|
||||
expect(slugFrom(run(SLUG_BIN, [], project, home).stdout)).toBe(initialSlug);
|
||||
} finally {
|
||||
fs.rmSync(home, { recursive: true, force: true });
|
||||
fs.rmSync(root, { recursive: true, force: true });
|
||||
}
|
||||
});
|
||||
}
|
||||
|
||||
test('still treats a workspace without origin as unknown', () => {
|
||||
const home = fs.mkdtempSync(path.join(os.tmpdir(), 'grepo-home-'));
|
||||
const project = fs.mkdtempSync(path.join(os.tmpdir(), 'grepo-local-only-'));
|
||||
try {
|
||||
git(['init', '-b', 'main'], project, home);
|
||||
const mode = run(REPO_MODE_BIN, [], project, home);
|
||||
expect(mode.status, mode.stderr).toBe(0);
|
||||
expect(mode.stdout.trim()).toBe('REPO_MODE=unknown');
|
||||
expect(fs.existsSync(path.join(home, '.gstack', 'projects'))).toBe(false);
|
||||
} finally {
|
||||
fs.rmSync(home, { recursive: true, force: true });
|
||||
fs.rmSync(project, { recursive: true, force: true });
|
||||
}
|
||||
});
|
||||
|
||||
test('subdir invocation still writes under the git-root cached slug', () => {
|
||||
const home = fs.mkdtempSync(path.join(os.tmpdir(), 'grepo-home-'));
|
||||
const root = fs.mkdtempSync(path.join(os.tmpdir(), 'grepo-subdir-'));
|
||||
const project = path.join(root, 'nested-history');
|
||||
const nested = path.join(project, 'packages', 'app');
|
||||
try {
|
||||
fs.mkdirSync(nested, { recursive: true });
|
||||
git(['init', '-b', 'main'], project, home);
|
||||
git(['config', 'user.name', 'gstack regression test'], project, home);
|
||||
git(['config', 'user.email', 'test@gstack.dev'], project, home);
|
||||
|
||||
const initialSlug = slugFrom(run(SLUG_BIN, [], project, home).stdout);
|
||||
expect(initialSlug).toBe('nested-history');
|
||||
|
||||
for (let commit = 1; commit <= 5; commit += 1) {
|
||||
fs.writeFileSync(path.join(project, `commit-${commit}.txt`), String(commit));
|
||||
git(['add', '.'], project, home);
|
||||
git(['commit', '-m', `commit ${commit}`], project, home);
|
||||
}
|
||||
git(['remote', 'add', 'origin', 'https://github.com/facebook/react.git'], project, home);
|
||||
|
||||
const mode = run(REPO_MODE_BIN, [], nested, home);
|
||||
expect(mode.status, mode.stderr).toBe(0);
|
||||
expect(mode.stdout.trim()).toBe('REPO_MODE=solo');
|
||||
|
||||
const projectsDir = path.join(home, '.gstack', 'projects');
|
||||
expect(fs.existsSync(path.join(projectsDir, initialSlug, 'repo-mode.json'))).toBe(true);
|
||||
expect(fs.readdirSync(projectsDir)).toEqual([initialSlug]);
|
||||
} finally {
|
||||
fs.rmSync(home, { recursive: true, force: true });
|
||||
fs.rmSync(root, { recursive: true, force: true });
|
||||
}
|
||||
});
|
||||
|
||||
test('sanitizes unusual origin URLs via gstack-slug before mkdir', () => {
|
||||
const home = fs.mkdtempSync(path.join(os.tmpdir(), 'grepo-home-'));
|
||||
const project = fs.mkdtempSync(path.join(os.tmpdir(), 'grepo-weird-url-'));
|
||||
try {
|
||||
git(['init', '-b', 'main'], project, home);
|
||||
git(['config', 'user.name', 'gstack regression test'], project, home);
|
||||
git(['config', 'user.email', 'test@gstack.dev'], project, home);
|
||||
|
||||
for (let commit = 1; commit <= 5; commit += 1) {
|
||||
fs.writeFileSync(path.join(project, `commit-${commit}.txt`), String(commit));
|
||||
git(['add', '.'], project, home);
|
||||
git(['commit', '-m', `commit ${commit}`], project, home);
|
||||
}
|
||||
// Free-form remote URL with characters outside [a-zA-Z0-9._-]
|
||||
git(['remote', 'add', 'origin', 'https://example.com/org/my repo (v2).git'], project, home);
|
||||
|
||||
const mode = run(REPO_MODE_BIN, [], project, home);
|
||||
expect(mode.status, mode.stderr).toBe(0);
|
||||
expect(mode.stdout.trim()).toBe('REPO_MODE=solo');
|
||||
|
||||
const projects = fs.readdirSync(path.join(home, '.gstack', 'projects'));
|
||||
expect(projects).toHaveLength(1);
|
||||
expect(projects[0]).toMatch(/^[a-zA-Z0-9._-]+$/);
|
||||
expect(projects[0]).not.toMatch(/[ ()]/);
|
||||
} finally {
|
||||
fs.rmSync(home, { recursive: true, force: true });
|
||||
fs.rmSync(project, { recursive: true, force: true });
|
||||
}
|
||||
});
|
||||
});
|
||||
@@ -8,7 +8,8 @@
|
||||
* Key differences from Codex session-runner:
|
||||
* - Uses `gemini -p` instead of `codex exec`
|
||||
* - Output is NDJSON with event types: init, message, tool_use, tool_result, result
|
||||
* - Uses `--output-format stream-json --yolo` instead of `--json -s read-only`
|
||||
* - Uses `--output-format stream-json --yolo --skip-trust` instead of `--json -s read-only`
|
||||
* (`--skip-trust` required for headless/untrusted cwds; see gemini trusted-folders docs)
|
||||
* - No temp HOME needed — Gemini discovers skills from `.agents/skills/` in cwd
|
||||
* - Message events are streamed with `delta: true` — must concatenate
|
||||
*/
|
||||
@@ -121,7 +122,9 @@ export async function runGeminiSkill(opts: {
|
||||
}
|
||||
|
||||
// Build gemini command
|
||||
const args = ['-p', prompt, '--output-format', 'stream-json', '--yolo'];
|
||||
// --skip-trust: headless/CI and temp cwds aren't in ~/.gemini/trustedFolders.json;
|
||||
// without it gemini exits FatalUntrustedWorkspaceError before any model call.
|
||||
const args = ['-p', prompt, '--output-format', 'stream-json', '--yolo', '--skip-trust'];
|
||||
|
||||
// Spawn gemini — uses real HOME for auth (~/.gemini; HOME is allowlisted),
|
||||
// cwd for skill discovery. Hermetic scrub with gemini's auth surface
|
||||
|
||||
@@ -0,0 +1,138 @@
|
||||
import { describe, test, expect } from 'bun:test';
|
||||
import { parseGeminiStreamJson, resultFromGeminiStream } from './gemini';
|
||||
|
||||
// Current CLI shape (gemini ≥ ~0.11 / stream-json): content + role, tokens under result.stats
|
||||
// Documented at https://geminicli.com/docs/cli/headless/ and gemini-cli PR #10883.
|
||||
const CURRENT_CLI_FIXTURE = [
|
||||
'{"type":"init","timestamp":"2026-03-20T15:14:46.455Z","session_id":"test-session-123","model":"gemini-2.5-pro"}',
|
||||
'{"type":"message","role":"user","content":"Reply with exactly the word: PONG"}',
|
||||
'{"type":"message","role":"assistant","content":"PONG","delta":true}',
|
||||
'{"type":"result","status":"success","stats":{"input_tokens":24946,"output_tokens":32,"total_tokens":24978}}',
|
||||
].join('\n');
|
||||
|
||||
// Legacy shape: text field, tokens under result.usage
|
||||
const LEGACY_FIXTURE = [
|
||||
'{"type":"message","text":"hello"}',
|
||||
'{"type":"tool_use","name":"run_shell_command"}',
|
||||
'{"type":"result","usage":{"input_token_count":100,"output_token_count":5},"model":"gemini-2.5-flash"}',
|
||||
].join('\n');
|
||||
|
||||
describe('parseGeminiStreamJson', () => {
|
||||
test('current CLI: reads assistant content, ignores user echo, stats tokens, init model', () => {
|
||||
const parsed = parseGeminiStreamJson(CURRENT_CLI_FIXTURE);
|
||||
expect(parsed.output).toBe('PONG');
|
||||
expect(parsed.tokens.input).toBe(24946);
|
||||
expect(parsed.tokens.output).toBe(32);
|
||||
expect(parsed.modelUsed).toBe('gemini-2.5-pro');
|
||||
expect(parsed.toolCalls).toBe(0);
|
||||
});
|
||||
|
||||
test('current CLI: user role content must not concat into output', () => {
|
||||
const raw = [
|
||||
'{"type":"message","role":"user","content":"PROMPT ECHO"}',
|
||||
'{"type":"message","role":"assistant","content":"ok","delta":true}',
|
||||
].join('\n');
|
||||
const parsed = parseGeminiStreamJson(raw);
|
||||
expect(parsed.output).toBe('ok');
|
||||
expect(parsed.output).not.toContain('PROMPT ECHO');
|
||||
});
|
||||
|
||||
test('legacy: still accepts text + usage.input_token_count', () => {
|
||||
const parsed = parseGeminiStreamJson(LEGACY_FIXTURE);
|
||||
expect(parsed.output).toBe('hello');
|
||||
expect(parsed.tokens.input).toBe(100);
|
||||
expect(parsed.tokens.output).toBe(5);
|
||||
expect(parsed.toolCalls).toBe(1);
|
||||
expect(parsed.modelUsed).toBe('gemini-2.5-flash');
|
||||
});
|
||||
|
||||
test('legacy text with role:user is ignored', () => {
|
||||
const raw = [
|
||||
'{"type":"message","role":"user","text":"echo"}',
|
||||
'{"type":"message","role":"assistant","text":"kept"}',
|
||||
].join('\n');
|
||||
const parsed = parseGeminiStreamJson(raw);
|
||||
expect(parsed.output).toBe('kept');
|
||||
});
|
||||
|
||||
test('concatenates multiple assistant content deltas', () => {
|
||||
const raw = [
|
||||
'{"type":"message","role":"assistant","content":"A","delta":true}',
|
||||
'{"type":"message","role":"assistant","content":"B","delta":true}',
|
||||
].join('\n');
|
||||
expect(parseGeminiStreamJson(raw).output).toBe('AB');
|
||||
});
|
||||
|
||||
test('skips malformed lines without throwing', () => {
|
||||
const raw = [
|
||||
'{"type":"init","model":"m1"}',
|
||||
'not json',
|
||||
'{"type":"message","role":"assistant","content":"x","delta":true}',
|
||||
'{incomplete',
|
||||
'{"type":"result","stats":{"input_tokens":1,"output_tokens":2}}',
|
||||
].join('\n');
|
||||
const parsed = parseGeminiStreamJson(raw);
|
||||
expect(parsed.output).toBe('x');
|
||||
expect(parsed.tokens).toEqual({ input: 1, output: 2 });
|
||||
expect(parsed.modelUsed).toBe('m1');
|
||||
});
|
||||
|
||||
test('empty / whitespace-only input yields empty parse (no throw)', () => {
|
||||
const parsed = parseGeminiStreamJson('');
|
||||
expect(parsed.output).toBe('');
|
||||
expect(parsed.tokens).toEqual({ input: 0, output: 0 });
|
||||
expect(parsed.toolCalls).toBe(0);
|
||||
expect(parsed.modelUsed).toBeUndefined();
|
||||
});
|
||||
|
||||
test('result.model overrides init.model when both present', () => {
|
||||
const raw = [
|
||||
'{"type":"init","model":"from-init"}',
|
||||
'{"type":"message","role":"assistant","content":"hi"}',
|
||||
'{"type":"result","model":"from-result","stats":{"input_tokens":1,"output_tokens":1}}',
|
||||
].join('\n');
|
||||
expect(parseGeminiStreamJson(raw).modelUsed).toBe('from-result');
|
||||
});
|
||||
});
|
||||
|
||||
describe('resultFromGeminiStream (adapter post-CLI e2e)', () => {
|
||||
test('current CLI fixture → success row with PONG + tokens (not $0)', () => {
|
||||
const result = resultFromGeminiStream(CURRENT_CLI_FIXTURE, { durationMs: 42 });
|
||||
expect(result.error).toBeUndefined();
|
||||
expect(result.output).toBe('PONG');
|
||||
expect(result.tokens.input).toBe(24946);
|
||||
expect(result.tokens.output).toBe(32);
|
||||
expect(result.modelUsed).toBe('gemini-2.5-pro');
|
||||
expect(result.durationMs).toBe(42);
|
||||
// Cost signal: non-zero tokens means estimateCost will not be $0.
|
||||
expect(result.tokens.input + result.tokens.output).toBeGreaterThan(0);
|
||||
});
|
||||
|
||||
test('legacy text-only success still works', () => {
|
||||
const result = resultFromGeminiStream(LEGACY_FIXTURE, { durationMs: 10 });
|
||||
expect(result.error).toBeUndefined();
|
||||
expect(result.output).toBe('hello');
|
||||
expect(result.toolCalls).toBe(1);
|
||||
});
|
||||
|
||||
test('empty exit-0 stream → error row, never silent $0 success (#2159)', () => {
|
||||
const emptySuccess = [
|
||||
'{"type":"init","model":"gemini-2.5-pro"}',
|
||||
'{"type":"message","role":"user","content":"hi"}',
|
||||
'{"type":"result","status":"success","stats":{"input_tokens":0,"output_tokens":0}}',
|
||||
].join('\n');
|
||||
const result = resultFromGeminiStream(emptySuccess, { durationMs: 5 });
|
||||
expect(result.error).toBeDefined();
|
||||
expect(result.error!.code).toBe('unknown');
|
||||
expect(result.error!.reason).toContain('empty output');
|
||||
expect(result.output).toBe('');
|
||||
expect(result.modelUsed).toBe('gemini-2.5-pro');
|
||||
});
|
||||
|
||||
test('pre-fix bug shape (text field missing, only content) would have been empty — now succeeds', () => {
|
||||
// This is the exact failure mode from #2159: content present, no text.
|
||||
const result = resultFromGeminiStream(CURRENT_CLI_FIXTURE);
|
||||
expect(result.error).toBeUndefined();
|
||||
expect(result.output).toBe('PONG');
|
||||
});
|
||||
});
|
||||
@@ -5,13 +5,108 @@ import * as fs from 'fs';
|
||||
import * as path from 'path';
|
||||
import * as os from 'os';
|
||||
|
||||
export type GeminiStreamParse = {
|
||||
output: string;
|
||||
tokens: { input: number; output: number };
|
||||
toolCalls: number;
|
||||
modelUsed?: string;
|
||||
};
|
||||
|
||||
/**
|
||||
* Parse gemini NDJSON stream events (exported for unit tests).
|
||||
*
|
||||
* Current CLI (`--output-format stream-json`) emits:
|
||||
* init → model
|
||||
* message { role, content, delta? } → concat assistant content
|
||||
* tool_use → increment toolCalls
|
||||
* result { stats: { input_tokens, output_tokens } } → tokens
|
||||
*
|
||||
* Legacy shape (still accepted):
|
||||
* message { text } → concat text
|
||||
* result { usage: { input_token_count, output_token_count } } → tokens
|
||||
*/
|
||||
export function parseGeminiStreamJson(raw: string): GeminiStreamParse {
|
||||
let output = '';
|
||||
let input = 0;
|
||||
let out = 0;
|
||||
let toolCalls = 0;
|
||||
let modelUsed: string | undefined;
|
||||
for (const line of raw.split('\n')) {
|
||||
const s = line.trim();
|
||||
if (!s) continue;
|
||||
try {
|
||||
const obj = JSON.parse(s);
|
||||
if (obj.type === 'init') {
|
||||
if (typeof obj.model === 'string' && obj.model) modelUsed = obj.model;
|
||||
} else if (obj.type === 'message') {
|
||||
// Current CLI: content + role. Role guard is required — the CLI echoes
|
||||
// the user prompt as role:'user', which must not land in output.
|
||||
if (obj.role === 'assistant' && typeof obj.content === 'string') {
|
||||
output += obj.content;
|
||||
} else if (typeof obj.text === 'string' && obj.role !== 'user') {
|
||||
// Legacy text field (no role, or assistant).
|
||||
output += obj.text;
|
||||
}
|
||||
} else if (obj.type === 'tool_use') {
|
||||
toolCalls += 1;
|
||||
} else if (obj.type === 'result') {
|
||||
const u = obj.usage ?? obj.stats ?? {};
|
||||
input += u.input_token_count ?? u.input_tokens ?? u.prompt_tokens ?? 0;
|
||||
out += u.output_token_count ?? u.output_tokens ?? u.completion_tokens ?? 0;
|
||||
if (typeof obj.model === 'string' && obj.model) modelUsed = obj.model;
|
||||
}
|
||||
} catch {
|
||||
// skip malformed lines
|
||||
}
|
||||
}
|
||||
return { output, tokens: { input, output: out }, toolCalls, modelUsed };
|
||||
}
|
||||
|
||||
/**
|
||||
* Map a raw stream-json dump to a RunResult, including the empty-success
|
||||
* hardening from #2159. Exported so adapter e2e can exercise the full
|
||||
* post-CLI path without a live gemini binary.
|
||||
*/
|
||||
export function resultFromGeminiStream(
|
||||
raw: string,
|
||||
opts: { model?: string; durationMs?: number } = {},
|
||||
): RunResult {
|
||||
const parsed = parseGeminiStreamJson(raw);
|
||||
const modelUsed = parsed.modelUsed || opts.model || 'gemini-2.5-pro';
|
||||
const durationMs = opts.durationMs ?? 0;
|
||||
if (!parsed.output.trim()) {
|
||||
return {
|
||||
output: '',
|
||||
tokens: { input: 0, output: 0 },
|
||||
durationMs,
|
||||
toolCalls: 0,
|
||||
modelUsed,
|
||||
error: { code: 'unknown', reason: 'empty output from gemini CLI (exit 0)' },
|
||||
};
|
||||
}
|
||||
return {
|
||||
output: parsed.output,
|
||||
tokens: parsed.tokens,
|
||||
durationMs,
|
||||
toolCalls: parsed.toolCalls,
|
||||
modelUsed,
|
||||
};
|
||||
}
|
||||
|
||||
/**
|
||||
* Gemini adapter — wraps the `gemini` CLI.
|
||||
*
|
||||
* Gemini CLI auth comes from either ~/.config/gemini/ or GOOGLE_API_KEY. Output
|
||||
* format is NDJSON with `message`/`tool_use`/`result` events when `--output-format
|
||||
* stream-json` is requested. This adapter uses a single-response form for simplicity
|
||||
* in benchmarks; richer streaming lives in gemini-session-runner.ts.
|
||||
* Auth: GEMINI_API_KEY / GOOGLE_API_KEY (preferred), or ~/.gemini oauth.
|
||||
* Personal OAuth free-tier is no longer supported by gemini CLI — use an
|
||||
* AI Studio API key. Antigravity is a separate product/quota path.
|
||||
*
|
||||
* Headless flags always passed:
|
||||
* --output-format stream-json — NDJSON events (message/tool_use/result)
|
||||
* --yolo — auto-approve tools (non-interactive)
|
||||
* --skip-trust — trust cwd for this session; required when
|
||||
* workdir is a temp/untrusted folder (benchmarks
|
||||
* use mkdtemp). Without it headless gemini exits
|
||||
* before calling the model.
|
||||
*/
|
||||
export class GeminiAdapter implements ProviderAdapter {
|
||||
readonly name = 'gemini';
|
||||
@@ -26,9 +121,14 @@ export class GeminiAdapter implements ProviderAdapter {
|
||||
const newCfgDir = path.join(os.homedir(), '.gemini');
|
||||
const newOauth = path.join(newCfgDir, 'oauth_creds.json');
|
||||
const hasCfg = fs.existsSync(legacyCfgDir) || fs.existsSync(newOauth);
|
||||
const hasKey = !!process.env.GOOGLE_API_KEY;
|
||||
// CLI accepts either name; Google AI Studio keys are usually GEMINI_API_KEY.
|
||||
const hasKey = !!(process.env.GOOGLE_API_KEY || process.env.GEMINI_API_KEY);
|
||||
if (!hasCfg && !hasKey) {
|
||||
return { ok: false, reason: 'No Gemini auth found. Log in via `gemini login` or export GOOGLE_API_KEY.' };
|
||||
return {
|
||||
ok: false,
|
||||
reason:
|
||||
'No Gemini auth found. Export GEMINI_API_KEY (or GOOGLE_API_KEY) from https://aistudio.google.com/app/apikey — personal OAuth free-tier is no longer supported by gemini CLI.',
|
||||
};
|
||||
}
|
||||
return { ok: true };
|
||||
}
|
||||
@@ -36,8 +136,10 @@ export class GeminiAdapter implements ProviderAdapter {
|
||||
async run(opts: RunOpts): Promise<RunResult> {
|
||||
const start = Date.now();
|
||||
// Default to --yolo (non-interactive) and stream-json output so we can parse
|
||||
// tokens + tool calls. Callers can override via extraArgs.
|
||||
const args = ['-p', opts.prompt, '--output-format', 'stream-json', '--yolo'];
|
||||
// tokens + tool calls. --skip-trust is required for headless/temp workdirs
|
||||
// (gemini CLI otherwise exits: "not running in a trusted directory").
|
||||
// Callers can override via extraArgs.
|
||||
const args = ['-p', opts.prompt, '--output-format', 'stream-json', '--yolo', '--skip-trust'];
|
||||
if (opts.model) args.push('--model', opts.model);
|
||||
if (opts.extraArgs) args.push(...opts.extraArgs);
|
||||
|
||||
@@ -47,15 +149,15 @@ export class GeminiAdapter implements ProviderAdapter {
|
||||
timeout: opts.timeoutMs,
|
||||
encoding: 'utf-8',
|
||||
maxBuffer: 32 * 1024 * 1024,
|
||||
env: {
|
||||
...process.env,
|
||||
// Prefer GEMINI_API_KEY when only that is set (CLI reads both).
|
||||
...(process.env.GEMINI_API_KEY && !process.env.GOOGLE_API_KEY
|
||||
? { GOOGLE_API_KEY: process.env.GEMINI_API_KEY }
|
||||
: {}),
|
||||
},
|
||||
});
|
||||
const parsed = this.parseStreamJson(out);
|
||||
return {
|
||||
output: parsed.output,
|
||||
tokens: parsed.tokens,
|
||||
durationMs: Date.now() - start,
|
||||
toolCalls: parsed.toolCalls,
|
||||
modelUsed: parsed.modelUsed || opts.model || 'gemini-2.5-pro',
|
||||
};
|
||||
return resultFromGeminiStream(out, { model: opts.model, durationMs: Date.now() - start });
|
||||
} catch (err: unknown) {
|
||||
const durationMs = Date.now() - start;
|
||||
const e = err as { code?: string; stderr?: Buffer; signal?: string; message?: string };
|
||||
@@ -63,7 +165,7 @@ export class GeminiAdapter implements ProviderAdapter {
|
||||
if (e.signal === 'SIGTERM' || e.code === 'ETIMEDOUT') {
|
||||
return this.emptyResult(durationMs, { code: 'timeout', reason: `exceeded ${opts.timeoutMs}ms` }, opts.model);
|
||||
}
|
||||
if (/unauthorized|auth|login|api key/i.test(stderr)) {
|
||||
if (/unauthorized|auth|login|api key|ineligibletier|no longer supported/i.test(stderr)) {
|
||||
return this.emptyResult(durationMs, { code: 'auth', reason: stderr.slice(0, 400) }, opts.model);
|
||||
}
|
||||
if (/rate[- ]?limit|429|quota/i.test(stderr)) {
|
||||
@@ -77,41 +179,6 @@ export class GeminiAdapter implements ProviderAdapter {
|
||||
return estimateCostUsd(tokens, model ?? 'gemini-2.5-pro');
|
||||
}
|
||||
|
||||
/**
|
||||
* Parse gemini NDJSON stream events:
|
||||
* init → session id (discarded here)
|
||||
* message { delta: true, text } → concat to output
|
||||
* tool_use { name } → increment toolCalls
|
||||
* result { usage: { input_token_count, output_token_count } } → tokens
|
||||
*/
|
||||
private parseStreamJson(raw: string): { output: string; tokens: { input: number; output: number }; toolCalls: number; modelUsed?: string } {
|
||||
let output = '';
|
||||
let input = 0;
|
||||
let out = 0;
|
||||
let toolCalls = 0;
|
||||
let modelUsed: string | undefined;
|
||||
for (const line of raw.split('\n')) {
|
||||
const s = line.trim();
|
||||
if (!s) continue;
|
||||
try {
|
||||
const obj = JSON.parse(s);
|
||||
if (obj.type === 'message' && typeof obj.text === 'string') {
|
||||
output += obj.text;
|
||||
} else if (obj.type === 'tool_use') {
|
||||
toolCalls += 1;
|
||||
} else if (obj.type === 'result') {
|
||||
const u = obj.usage ?? {};
|
||||
input += u.input_token_count ?? u.prompt_tokens ?? 0;
|
||||
out += u.output_token_count ?? u.completion_tokens ?? 0;
|
||||
if (obj.model) modelUsed = obj.model;
|
||||
}
|
||||
} catch {
|
||||
// skip malformed lines
|
||||
}
|
||||
}
|
||||
return { output, tokens: { input, output: out }, toolCalls, modelUsed };
|
||||
}
|
||||
|
||||
private emptyResult(durationMs: number, error: RunResult['error'], model?: string): RunResult {
|
||||
return {
|
||||
output: '',
|
||||
|
||||
@@ -129,19 +129,15 @@ describeIfEvals('multi-provider benchmark adapters (live)', () => {
|
||||
if (result.error) {
|
||||
throw new Error(`gemini errored: ${result.error.code} — ${result.error.reason}`);
|
||||
}
|
||||
// Gemini CLI occasionally returns empty output even on successful runs
|
||||
// (model returned content the CLI parser missed, intermittent stream issues).
|
||||
// We assert the adapter ran end-to-end without erroring and reports a non-
|
||||
// empty token count instead of grepping the literal "ok" — that string
|
||||
// assertion was too brittle for a smoke that's really about "did the
|
||||
// adapter wire up and the run terminate successfully?"
|
||||
expect(typeof result.output).toBe('string');
|
||||
// Gemini CLI sometimes returns 0 tokens in the result event (older responses);
|
||||
// assert non-negative instead of strictly positive.
|
||||
expect(result.tokens.input).toBeGreaterThanOrEqual(0);
|
||||
expect(result.tokens.output).toBeGreaterThanOrEqual(0);
|
||||
// Adapter must never report empty-success (#2159). After content/stats
|
||||
// parsing, a healthy run has non-empty assistant text + token counts.
|
||||
expect(result.output.trim().length).toBeGreaterThan(0);
|
||||
expect(result.output.toLowerCase()).toContain('ok');
|
||||
expect(result.tokens.input).toBeGreaterThan(0);
|
||||
expect(result.tokens.output).toBeGreaterThan(0);
|
||||
expect(result.durationMs).toBeGreaterThan(0);
|
||||
expect(typeof result.modelUsed).toBe('string');
|
||||
expect(result.modelUsed.length).toBeGreaterThan(0);
|
||||
}, 150_000);
|
||||
|
||||
test('timeout error surfaces as error.code=timeout (no exception)', async () => {
|
||||
|
||||
@@ -8,7 +8,7 @@
|
||||
* 3. sha8($(git config user.email))
|
||||
* 4. anonymous-<sha8(hostname)>
|
||||
*
|
||||
* Result is persisted under user_slug_at_<endpoint-hash> for stability.
|
||||
* Result is persisted under user_slug_at_<endpoint-id> for stability.
|
||||
* Test isolation via GSTACK_HOME and HOME env overrides.
|
||||
*
|
||||
* Gate-tier, free, ~50ms.
|
||||
@@ -87,12 +87,12 @@ describe('resolve-user-slug fallback chain', () => {
|
||||
expect(slug).toMatch(/^(email-|anonymous-)[a-f0-9]+$|^[a-zA-Z0-9-]+$/);
|
||||
});
|
||||
|
||||
test('persists resolution to user_slug_at_<hash> on first call', () => {
|
||||
test('persists resolution to user_slug_at_<endpoint-id> on first call', () => {
|
||||
runConfig(['resolve-user-slug'], { GSTACK_HOME: TMP_HOME, USER: 'persisttest' });
|
||||
const configFile = join(TMP_HOME, 'config.yaml');
|
||||
expect(existsSync(configFile)).toBe(true);
|
||||
const content = readFileSync(configFile, 'utf-8');
|
||||
expect(content).toMatch(/^user_slug_at_[a-f0-9]+:\s+persisttest/m);
|
||||
expect(content).toMatch(/^user_slug_at_(local|[a-f0-9]{8}|[a-f0-9]{16}):\s+persisttest/m);
|
||||
});
|
||||
|
||||
test('subsequent calls return same slug (stable across sessions)', () => {
|
||||
@@ -104,7 +104,7 @@ describe('resolve-user-slug fallback chain', () => {
|
||||
});
|
||||
});
|
||||
|
||||
describe('brain_trust_policy@<hash> namespace', () => {
|
||||
describe('brain_trust_policy@<endpoint-id> namespace', () => {
|
||||
test('default value is "unset"', () => {
|
||||
const result = runConfig(['get', 'brain_trust_policy@deadbeef'], { GSTACK_HOME: TMP_HOME });
|
||||
expect(result.status).toBe(0);
|
||||
@@ -158,4 +158,10 @@ describe('key validation', () => {
|
||||
const result = runConfig(['get', 'brain_trust_policy@abc123ff'], { GSTACK_HOME: TMP_HOME });
|
||||
expect(result.status).toBe(0);
|
||||
});
|
||||
|
||||
test('accepts @local suffix on key', () => {
|
||||
const result = runConfig(['get', 'brain_trust_policy@local'], { GSTACK_HOME: TMP_HOME });
|
||||
expect(result.status).toBe(0);
|
||||
expect(result.stdout).toBe('unset');
|
||||
});
|
||||
});
|
||||
|
||||
Reference in New Issue
Block a user