feat(mosaic): WI-2 mutator-class guard for directive-freshness (#837)
WI-2: mutator-class guard for the compaction directive-freshness mechanism — command-position parser (prefix x var-indirection unified) + primitive-anchored invariant backstop + all-tools-hook fail-close. Parser-complete on principle: realistic evasion matrix RED-regression-covered, residual exotic evasions proven backstop-caught (B-tests), lens-convergence reached. closes #829
This commit was merged in pull request #837.
This commit is contained in:
667
packages/mosaic/src/mutator-gate/mutator-gate.acceptance.spec.ts
Normal file
667
packages/mosaic/src/mutator-gate/mutator-gate.acceptance.spec.ts
Normal file
@@ -0,0 +1,667 @@
|
||||
import { chmod, mkdir, mkdtemp, readFile, rm, writeFile } from 'node:fs/promises';
|
||||
import { createConnection } from 'node:net';
|
||||
import { tmpdir } from 'node:os';
|
||||
import { join } from 'node:path';
|
||||
import { spawn, spawnSync, type ChildProcess } from 'node:child_process';
|
||||
|
||||
import { afterEach, describe, expect, test } from 'vitest';
|
||||
|
||||
import { launchClaudex, type ClaudexHarnessAdapter } from '../commands/claudex.js';
|
||||
|
||||
interface BrokerReply {
|
||||
ok: boolean;
|
||||
code?: string;
|
||||
decision?: 'allow' | 'deny';
|
||||
state?: 'UNVERIFIED' | 'PENDING_VERIFICATION' | 'VERIFIED';
|
||||
session_id?: string;
|
||||
promotion_token?: string;
|
||||
}
|
||||
|
||||
interface BrokerPaths {
|
||||
socket: string;
|
||||
}
|
||||
|
||||
const frameworkRoot = new URL('../../framework/', import.meta.url).pathname;
|
||||
const daemonPath = join(frameworkRoot, 'tools/lease-broker/daemon.py');
|
||||
const gatePath = join(frameworkRoot, 'tools/lease-broker/mutator-gate.py');
|
||||
const launchGuardPath = join(frameworkRoot, 'tools/lease-broker/check-runtime-launches.py');
|
||||
const launcherPath = join(frameworkRoot, 'tools/lease-broker/launch-runtime.py');
|
||||
const claudeSettingsPath = join(frameworkRoot, 'runtime/claude/settings.json');
|
||||
const piExtensionPath = join(frameworkRoot, 'runtime/pi/mosaic-extension.ts');
|
||||
const prdyInitPath = join(frameworkRoot, 'tools/prdy/prdy-init.sh');
|
||||
const prdyUpdatePath = join(frameworkRoot, 'tools/prdy/prdy-update.sh');
|
||||
const remediationHandlerPath = join(frameworkRoot, 'tools/qa/remediation-hook-handler.sh');
|
||||
const children: ChildProcess[] = [];
|
||||
const temporaryRoots: string[] = [];
|
||||
|
||||
const binding = (compaction_epoch = 1) => ({
|
||||
compaction_epoch,
|
||||
request_epoch: 0,
|
||||
h_source: 'a'.repeat(64),
|
||||
h_payload: 'b'.repeat(64),
|
||||
schema_version: 1,
|
||||
});
|
||||
|
||||
async function request(socketPath: string, requestValue: object): Promise<BrokerReply> {
|
||||
return await new Promise<BrokerReply>((resolve, reject) => {
|
||||
const socket = createConnection(socketPath);
|
||||
let response = '';
|
||||
socket.setEncoding('utf8');
|
||||
socket.once('error', reject);
|
||||
socket.on('data', (chunk: string) => {
|
||||
response += chunk;
|
||||
});
|
||||
socket.once('end', () => resolve(JSON.parse(response) as BrokerReply));
|
||||
socket.once('connect', () => socket.end(`${JSON.stringify(requestValue)}\n`));
|
||||
});
|
||||
}
|
||||
|
||||
async function startBroker(): Promise<BrokerPaths> {
|
||||
const root = await mkdtemp(join(tmpdir(), 'mosaic-mutator-gate-'));
|
||||
await chmod(root, 0o700);
|
||||
const socket = join(root, 'broker.sock');
|
||||
const child = spawn(
|
||||
'python3',
|
||||
[daemonPath, '--socket', socket, '--state', join(root, 'state.json')],
|
||||
{
|
||||
stdio: ['ignore', 'pipe', 'pipe'],
|
||||
},
|
||||
);
|
||||
children.push(child);
|
||||
await new Promise<void>((resolve, reject) => {
|
||||
let stderr = '';
|
||||
child.stderr?.setEncoding('utf8');
|
||||
child.stderr?.on('data', (chunk: string) => (stderr += chunk));
|
||||
child.once('error', reject);
|
||||
child.once('exit', (code: number | null) =>
|
||||
reject(new Error(`broker exited ${code}: ${stderr}`)),
|
||||
);
|
||||
child.stdout?.once('data', () => resolve());
|
||||
});
|
||||
return { socket };
|
||||
}
|
||||
|
||||
interface RuntimeLaunchEntry {
|
||||
name: string;
|
||||
script: string;
|
||||
prepare(root: string): Promise<string[]>;
|
||||
}
|
||||
|
||||
const runtimeLaunchEntries: RuntimeLaunchEntry[] = [
|
||||
{
|
||||
name: 'prdy-init',
|
||||
script: prdyInitPath,
|
||||
prepare: async (root) => ['--project', root, '--name', 'Gate Test'],
|
||||
},
|
||||
{
|
||||
name: 'prdy-update',
|
||||
script: prdyUpdatePath,
|
||||
prepare: async (root) => {
|
||||
await mkdir(join(root, 'docs'), { recursive: true });
|
||||
await writeFile(join(root, 'docs/PRD.md'), '# Existing PRD\n');
|
||||
return ['--project', root];
|
||||
},
|
||||
},
|
||||
{
|
||||
name: 'qa-remediation',
|
||||
script: remediationHandlerPath,
|
||||
prepare: async (root) => {
|
||||
const pending = join(root, 'reports/pending');
|
||||
await mkdir(pending, { recursive: true });
|
||||
const report = join(pending, 'gate_remediation_needed.md');
|
||||
await writeFile(report, '# remediation\n');
|
||||
return [report];
|
||||
},
|
||||
},
|
||||
];
|
||||
|
||||
async function runRuntimeLaunchEntry(entry: RuntimeLaunchEntry, socket: string) {
|
||||
const root = await mkdtemp(join(tmpdir(), `mosaic-${entry.name}-gate-`));
|
||||
temporaryRoots.push(root);
|
||||
const binDir = join(root, 'bin');
|
||||
await mkdir(binDir, { recursive: true });
|
||||
const fakeClaude = join(binDir, 'claude');
|
||||
await writeFile(
|
||||
fakeClaude,
|
||||
`#!/usr/bin/env python3
|
||||
import json
|
||||
import os
|
||||
import subprocess
|
||||
|
||||
session_id = os.environ.get("MOSAIC_LEASE_SESSION_ID", "")
|
||||
denied = subprocess.run(
|
||||
["python3", ${JSON.stringify(gatePath)}, "--runtime", "claude"],
|
||||
input=json.dumps({"tool_name": "Bash"}) + "\\n",
|
||||
text=True,
|
||||
capture_output=True,
|
||||
env=os.environ,
|
||||
).returncode == 2
|
||||
print("RUNTIME_PROBE=" + json.dumps({"session_id": session_id, "denied": denied}))
|
||||
raise SystemExit(0 if len(session_id) == 64 and denied else 1)
|
||||
`,
|
||||
{ mode: 0o700 },
|
||||
);
|
||||
await chmod(fakeClaude, 0o700);
|
||||
const args = await entry.prepare(root);
|
||||
return spawnSync('bash', [entry.script, ...args], {
|
||||
cwd: root,
|
||||
encoding: 'utf8',
|
||||
env: {
|
||||
...process.env,
|
||||
PATH: `${binDir}:${process.env.PATH ?? ''}`,
|
||||
MOSAIC_HOME: frameworkRoot,
|
||||
MOSAIC_PRDY_RUNTIME: 'claude',
|
||||
MOSAIC_LEASE_BROKER_SOCKET: socket,
|
||||
MOSAIC_RUNTIME_GENERATION: '1',
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
async function register(socket: string, runtime_generation = 1): Promise<string> {
|
||||
const reply = await request(socket, { action: 'register_anchor', runtime_generation });
|
||||
expect(reply.ok).toBe(true);
|
||||
expect(reply.session_id).toMatch(/^[a-f0-9]{64}$/);
|
||||
return reply.session_id!;
|
||||
}
|
||||
|
||||
async function beginVerification(
|
||||
socket: string,
|
||||
session_id: string,
|
||||
runtime: 'claude' | 'pi',
|
||||
runtime_generation = 1,
|
||||
ttl_seconds = 300,
|
||||
compactionEpoch = 1,
|
||||
): Promise<BrokerReply> {
|
||||
return await request(socket, {
|
||||
action: 'begin_verification',
|
||||
session_id,
|
||||
runtime_generation,
|
||||
runtime,
|
||||
ttl_seconds,
|
||||
binding: binding(compactionEpoch),
|
||||
});
|
||||
}
|
||||
|
||||
async function promote(
|
||||
socket: string,
|
||||
session_id: string,
|
||||
promotion_token: string,
|
||||
runtime_generation = 1,
|
||||
): Promise<BrokerReply> {
|
||||
return await request(socket, {
|
||||
action: 'promote_lease',
|
||||
session_id,
|
||||
runtime_generation,
|
||||
promotion_token,
|
||||
});
|
||||
}
|
||||
|
||||
async function authorize(
|
||||
socket: string,
|
||||
session_id: string,
|
||||
runtime: 'claude' | 'pi',
|
||||
tool_name: string,
|
||||
runtime_generation = 1,
|
||||
): Promise<BrokerReply> {
|
||||
return await request(socket, {
|
||||
action: 'authorize_tool',
|
||||
session_id,
|
||||
runtime_generation,
|
||||
runtime,
|
||||
tool_name,
|
||||
});
|
||||
}
|
||||
|
||||
function runRuntimeGate(
|
||||
socket: string,
|
||||
sessionId: string,
|
||||
runtime: 'claude' | 'pi',
|
||||
toolName: string,
|
||||
generation = 1,
|
||||
) {
|
||||
return spawnSync('python3', [gatePath, '--runtime', runtime], {
|
||||
input: `${JSON.stringify({ tool_name: toolName })}\n`,
|
||||
encoding: 'utf8',
|
||||
env: {
|
||||
...process.env,
|
||||
MOSAIC_LEASE_BROKER_SOCKET: socket,
|
||||
MOSAIC_LEASE_SESSION_ID: sessionId,
|
||||
MOSAIC_RUNTIME_GENERATION: String(generation),
|
||||
},
|
||||
});
|
||||
}
|
||||
|
||||
afterEach(async () => {
|
||||
for (const child of children.splice(0)) child.kill('SIGTERM');
|
||||
await Promise.all(
|
||||
temporaryRoots.splice(0).map((root) => rm(root, { recursive: true, force: true })),
|
||||
);
|
||||
});
|
||||
|
||||
describe('whole mutator-class lease gate', () => {
|
||||
test('revoke-first and promote-last structurally bracket mutator authority', async () => {
|
||||
const { socket } = await startBroker();
|
||||
const sessionId = await register(socket);
|
||||
|
||||
expect(await promote(socket, sessionId, 'c'.repeat(64))).toMatchObject({
|
||||
ok: false,
|
||||
code: 'INVALID_LEASE_TRANSITION',
|
||||
});
|
||||
expect(await authorize(socket, sessionId, 'claude', 'Bash')).toMatchObject({
|
||||
ok: false,
|
||||
code: 'MUTATOR_UNVERIFIED',
|
||||
decision: 'deny',
|
||||
});
|
||||
|
||||
const pending = await beginVerification(socket, sessionId, 'claude');
|
||||
expect(pending).toMatchObject({ ok: true, state: 'PENDING_VERIFICATION' });
|
||||
expect(pending.promotion_token).toMatch(/^[a-f0-9]{64}$/);
|
||||
expect(await authorize(socket, sessionId, 'claude', 'Write')).toMatchObject({
|
||||
ok: false,
|
||||
decision: 'deny',
|
||||
});
|
||||
|
||||
expect(await promote(socket, sessionId, pending.promotion_token!)).toMatchObject({
|
||||
ok: true,
|
||||
state: 'VERIFIED',
|
||||
});
|
||||
expect(await authorize(socket, sessionId, 'claude', 'Bash')).toMatchObject({
|
||||
ok: true,
|
||||
decision: 'allow',
|
||||
state: 'VERIFIED',
|
||||
});
|
||||
|
||||
const nextCycle = await beginVerification(socket, sessionId, 'claude', 1, 300, 2);
|
||||
expect(nextCycle).toMatchObject({ ok: true, state: 'PENDING_VERIFICATION' });
|
||||
expect(await authorize(socket, sessionId, 'claude', 'Edit')).toMatchObject({
|
||||
ok: false,
|
||||
decision: 'deny',
|
||||
});
|
||||
expect(await promote(socket, sessionId, pending.promotion_token!)).toMatchObject({
|
||||
ok: false,
|
||||
code: 'PROMOTION_TOKEN_MISMATCH',
|
||||
});
|
||||
});
|
||||
|
||||
test('non-dangerous parser residual is denied by the global all-tools hook without a lease', async () => {
|
||||
const root = await mkdtemp(join(tmpdir(), 'mosaic-parser-residual-'));
|
||||
temporaryRoots.push(root);
|
||||
const source = join(root, 'packages/probe/launch.sh');
|
||||
await mkdir(join(root, 'packages/probe'), { recursive: true });
|
||||
await writeFile(source, 'alias hidden_runtime=claude\nhidden_runtime -p x\n');
|
||||
|
||||
const parserResult = spawnSync('python3', [launchGuardPath, '--root', root, '--json'], {
|
||||
encoding: 'utf8',
|
||||
});
|
||||
expect(parserResult.status).toBe(0);
|
||||
expect(JSON.parse(parserResult.stdout)).toMatchObject({ gated: 0, total: 0 });
|
||||
|
||||
const settings = JSON.parse(await readFile(claudeSettingsPath, 'utf8')) as {
|
||||
hooks: { PreToolUse: Array<{ matcher: string; hooks: Array<{ command: string }> }> };
|
||||
};
|
||||
const allToolsHook = settings.hooks.PreToolUse.find((hook) => hook.matcher === '.*');
|
||||
expect(allToolsHook?.hooks[0]?.command).toContain('mutator-gate.py --runtime claude');
|
||||
|
||||
const environment = { ...process.env };
|
||||
delete environment['MOSAIC_LEASE_SESSION_ID'];
|
||||
delete environment['MOSAIC_LEASE_BROKER_SOCKET'];
|
||||
for (const toolName of ['Bash', 'Read', 'mcp__provider__custom']) {
|
||||
const gateResult = spawnSync('python3', [gatePath, '--runtime', 'claude'], {
|
||||
input: `${JSON.stringify({ tool_name: toolName })}\n`,
|
||||
encoding: 'utf8',
|
||||
env: environment,
|
||||
});
|
||||
expect(gateResult.status, toolName).toBe(2);
|
||||
expect(gateResult.stderr, toolName).toContain('GATE_UNAVAILABLE');
|
||||
}
|
||||
});
|
||||
|
||||
test('T-B raw and custom mutator tools are default-denied without shell parsing', async () => {
|
||||
const { socket } = await startBroker();
|
||||
const sessionId = await register(socket);
|
||||
const mutators: Array<['claude' | 'pi', string]> = [
|
||||
['claude', 'Bash'],
|
||||
['claude', 'Edit'],
|
||||
['claude', 'Write'],
|
||||
['claude', 'NotebookEdit'],
|
||||
['claude', 'mcp__provider__close_issue'],
|
||||
['pi', 'bash'],
|
||||
['pi', 'edit'],
|
||||
['pi', 'write'],
|
||||
['pi', 'deploy'],
|
||||
['pi', 'unknown_custom_tool'],
|
||||
];
|
||||
|
||||
for (const [runtime, toolName] of mutators) {
|
||||
expect(
|
||||
await authorize(socket, sessionId, runtime, toolName),
|
||||
`${runtime}:${toolName}`,
|
||||
).toMatchObject({
|
||||
ok: false,
|
||||
code: 'MUTATOR_UNVERIFIED',
|
||||
decision: 'deny',
|
||||
});
|
||||
}
|
||||
|
||||
for (const [runtime, toolName] of [
|
||||
['claude', 'Read'],
|
||||
['claude', 'Grep'],
|
||||
['pi', 'read'],
|
||||
['pi', 'grep'],
|
||||
['pi', 'mosaic_context_recover'],
|
||||
] as const) {
|
||||
expect(await authorize(socket, sessionId, runtime, toolName)).toMatchObject({
|
||||
ok: true,
|
||||
decision: 'allow',
|
||||
});
|
||||
}
|
||||
});
|
||||
|
||||
test('lease and tool validation failures remain fail-closed at the broker boundary', async () => {
|
||||
const { socket } = await startBroker();
|
||||
const sessionId = await register(socket);
|
||||
const baseRequest = {
|
||||
action: 'begin_verification',
|
||||
session_id: sessionId,
|
||||
runtime_generation: 1,
|
||||
runtime: 'claude',
|
||||
ttl_seconds: 300,
|
||||
binding: binding(),
|
||||
};
|
||||
|
||||
expect(await request(socket, { ...baseRequest, runtime: 'codex' })).toMatchObject({
|
||||
ok: false,
|
||||
code: 'INVALID_RUNTIME',
|
||||
});
|
||||
expect(
|
||||
await request(socket, { ...baseRequest, binding: { ...binding(), h_source: 'bad' } }),
|
||||
).toMatchObject({
|
||||
ok: false,
|
||||
code: 'INVALID_BINDING',
|
||||
});
|
||||
expect(await request(socket, { ...baseRequest, ttl_seconds: 0 })).toMatchObject({
|
||||
ok: false,
|
||||
code: 'INVALID_LEASE_TTL',
|
||||
});
|
||||
expect(
|
||||
await request(socket, {
|
||||
action: 'authorize_tool',
|
||||
session_id: sessionId,
|
||||
runtime_generation: 1,
|
||||
runtime: 'codex',
|
||||
tool_name: 'Read',
|
||||
}),
|
||||
).toMatchObject({ ok: false, code: 'INVALID_RUNTIME' });
|
||||
expect(
|
||||
await request(socket, {
|
||||
action: 'authorize_tool',
|
||||
session_id: sessionId,
|
||||
runtime_generation: 1,
|
||||
runtime: 'claude',
|
||||
tool_name: 'x'.repeat(257),
|
||||
}),
|
||||
).toMatchObject({ ok: false, code: 'INVALID_TOOL' });
|
||||
});
|
||||
|
||||
test('observer revocation and monotonic TTL expiry deny the next mutator', async () => {
|
||||
const { socket } = await startBroker();
|
||||
const sessionId = await register(socket);
|
||||
const pending = await beginVerification(socket, sessionId, 'claude', 1, 1);
|
||||
await promote(socket, sessionId, pending.promotion_token!);
|
||||
|
||||
expect(await authorize(socket, sessionId, 'claude', 'Bash')).toMatchObject({
|
||||
ok: true,
|
||||
decision: 'allow',
|
||||
});
|
||||
await new Promise((resolve) => setTimeout(resolve, 1_100));
|
||||
expect(await authorize(socket, sessionId, 'claude', 'Bash')).toMatchObject({
|
||||
ok: false,
|
||||
code: 'LEASE_EXPIRED',
|
||||
decision: 'deny',
|
||||
});
|
||||
|
||||
const refreshed = await beginVerification(socket, sessionId, 'claude', 1, 300, 2);
|
||||
await promote(socket, sessionId, refreshed.promotion_token!);
|
||||
expect(
|
||||
await request(socket, {
|
||||
action: 'revoke_lease',
|
||||
session_id: sessionId,
|
||||
runtime_generation: 1,
|
||||
reason: 'compaction_observer',
|
||||
}),
|
||||
).toMatchObject({ ok: true, state: 'UNVERIFIED' });
|
||||
expect(await authorize(socket, sessionId, 'claude', 'Write')).toMatchObject({
|
||||
ok: false,
|
||||
code: 'MUTATOR_UNVERIFIED',
|
||||
decision: 'deny',
|
||||
});
|
||||
});
|
||||
|
||||
test('runtime-generation replacement cannot inherit a verified lease', async () => {
|
||||
const { socket } = await startBroker();
|
||||
const sessionId = await register(socket);
|
||||
const pending = await beginVerification(socket, sessionId, 'pi');
|
||||
await promote(socket, sessionId, pending.promotion_token!);
|
||||
|
||||
expect(await authorize(socket, sessionId, 'pi', 'bash', 2)).toMatchObject({
|
||||
ok: false,
|
||||
code: 'MUTATOR_UNVERIFIED',
|
||||
decision: 'deny',
|
||||
});
|
||||
expect(await authorize(socket, sessionId, 'pi', 'bash', 1)).toMatchObject({
|
||||
ok: false,
|
||||
code: 'STALE_GENERATION',
|
||||
});
|
||||
});
|
||||
|
||||
test('runtime launcher anchors broker identity before exec and fails closed without broker', async () => {
|
||||
const { socket } = await startBroker();
|
||||
const probe = [
|
||||
'import json,os,socket',
|
||||
's=socket.socket(socket.AF_UNIX,socket.SOCK_STREAM)',
|
||||
"s.connect(os.environ['MOSAIC_LEASE_BROKER_SOCKET'])",
|
||||
"request={'action':'authorize_tool','session_id':os.environ['MOSAIC_LEASE_SESSION_ID'],'runtime_generation':int(os.environ['MOSAIC_RUNTIME_GENERATION']),'runtime':'claude','tool_name':'Read'}",
|
||||
"s.sendall((json.dumps(request)+'\\n').encode())",
|
||||
's.shutdown(socket.SHUT_WR)',
|
||||
"print(json.dumps({'session_id':os.environ['MOSAIC_LEASE_SESSION_ID'],'reply':json.loads(s.recv(65536))}))",
|
||||
].join(';');
|
||||
const launched = spawnSync(
|
||||
'python3',
|
||||
[launcherPath, '--runtime', 'claude', '--', 'python3', '-c', probe],
|
||||
{
|
||||
encoding: 'utf8',
|
||||
env: {
|
||||
...process.env,
|
||||
MOSAIC_LEASE_BROKER_SOCKET: socket,
|
||||
MOSAIC_RUNTIME_GENERATION: '1',
|
||||
},
|
||||
},
|
||||
);
|
||||
expect(launched.status, launched.stderr).toBe(0);
|
||||
expect(JSON.parse(launched.stdout)).toMatchObject({
|
||||
session_id: expect.stringMatching(/^[a-f0-9]{64}$/),
|
||||
reply: { ok: true, decision: 'allow' },
|
||||
});
|
||||
|
||||
const unavailable = spawnSync(
|
||||
'python3',
|
||||
[launcherPath, '--runtime', 'claude', '--', 'python3', '-c', "print('EXECUTED')"],
|
||||
{
|
||||
encoding: 'utf8',
|
||||
env: {
|
||||
...process.env,
|
||||
MOSAIC_LEASE_BROKER_SOCKET: join(tmpdir(), 'missing-mosaic-broker.sock'),
|
||||
MOSAIC_RUNTIME_GENERATION: '1',
|
||||
},
|
||||
},
|
||||
);
|
||||
expect(unavailable.status).not.toBe(0);
|
||||
expect(unavailable.stdout).not.toContain('EXECUTED');
|
||||
});
|
||||
|
||||
test.each(runtimeLaunchEntries)(
|
||||
'$name registers before launch, denies an unverified mutator, and fails closed without broker',
|
||||
async (entry) => {
|
||||
const missingSocket = join(tmpdir(), `missing-${entry.name}-${process.pid}.sock`);
|
||||
const unavailable = await runRuntimeLaunchEntry(entry, missingSocket);
|
||||
expect(unavailable.status).not.toBe(0);
|
||||
expect(`${unavailable.stdout}${unavailable.stderr}`).not.toContain('RUNTIME_PROBE=');
|
||||
|
||||
const { socket } = await startBroker();
|
||||
const launched = await runRuntimeLaunchEntry(entry, socket);
|
||||
expect(launched.status, launched.stderr).toBe(0);
|
||||
const match = /RUNTIME_PROBE=(\{[^\n]+\})/.exec(`${launched.stdout}${launched.stderr}`);
|
||||
expect(match).not.toBeNull();
|
||||
expect(JSON.parse(match![1]!)).toEqual({
|
||||
session_id: expect.stringMatching(/^[a-f0-9]{64}$/),
|
||||
denied: true,
|
||||
});
|
||||
},
|
||||
);
|
||||
|
||||
test.each([
|
||||
{ command: 'mosaic claudex', yolo: false },
|
||||
{ command: 'mosaic yolo claudex', yolo: true },
|
||||
])(
|
||||
'$command registers a broker anchor, installs the all-tools hook, and denies an unverified mutator',
|
||||
async ({ yolo }) => {
|
||||
const { socket } = await startBroker();
|
||||
const root = await mkdtemp(join(tmpdir(), 'mosaic-claudex-gate-'));
|
||||
temporaryRoots.push(root);
|
||||
const configDir = join(root, 'isolated-claude');
|
||||
const binDir = join(root, 'bin');
|
||||
await mkdir(configDir, { recursive: true });
|
||||
await mkdir(binDir, { recursive: true });
|
||||
|
||||
const fakeClaude = join(binDir, 'claude');
|
||||
const probe = `#!/usr/bin/env python3
|
||||
import json
|
||||
import os
|
||||
import subprocess
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
session_id = os.environ.get("MOSAIC_LEASE_SESSION_ID", "")
|
||||
settings_path = Path(os.environ["CLAUDE_CONFIG_DIR"]) / "settings.json"
|
||||
try:
|
||||
settings = json.loads(settings_path.read_text())
|
||||
except (OSError, json.JSONDecodeError):
|
||||
settings = {}
|
||||
pre_tool = settings.get("hooks", {}).get("PreToolUse", [])
|
||||
hook_present = any(
|
||||
item.get("matcher") == ".*" and any("mutator-gate.py" in hook.get("command", "") for hook in item.get("hooks", []))
|
||||
for item in pre_tool
|
||||
)
|
||||
denied = subprocess.run(
|
||||
["python3", ${JSON.stringify(gatePath)}, "--runtime", "claude"],
|
||||
input=json.dumps({"tool_name": "Bash"}) + "\\n",
|
||||
text=True,
|
||||
capture_output=True,
|
||||
env=os.environ,
|
||||
).returncode == 2
|
||||
is_yolo = "--dangerously-skip-permissions" in sys.argv[1:]
|
||||
result = {
|
||||
"session_id": session_id,
|
||||
"hook_present": hook_present,
|
||||
"denied": denied,
|
||||
"is_yolo": is_yolo,
|
||||
}
|
||||
print(json.dumps(result))
|
||||
raise SystemExit(0 if len(session_id) == 64 and hook_present and denied else 1)
|
||||
`;
|
||||
await writeFile(fakeClaude, probe, { mode: 0o700 });
|
||||
await chmod(fakeClaude, 0o700);
|
||||
|
||||
let execution: ReturnType<typeof spawnSync> | undefined;
|
||||
const run = (cmd: string, args: string[], env: NodeJS.ProcessEnv) => {
|
||||
execution = spawnSync(cmd, args, { encoding: 'utf8', env });
|
||||
};
|
||||
const adapter = {
|
||||
harnessPreflight: () => {},
|
||||
composePrompt: () => '# composed Claude contract',
|
||||
// Claudex exposes only the shared register-before-exec boundary.
|
||||
execLeaseGated: (args: string[], env: NodeJS.ProcessEnv, dangerous: boolean) =>
|
||||
run(
|
||||
'python3',
|
||||
[
|
||||
launcherPath,
|
||||
...(dangerous ? ['--dangerous'] : []),
|
||||
'--runtime',
|
||||
'claude',
|
||||
'--',
|
||||
'claude',
|
||||
...args,
|
||||
],
|
||||
env,
|
||||
),
|
||||
} as unknown as ClaudexHarnessAdapter;
|
||||
|
||||
await launchClaudex([], yolo, adapter, {
|
||||
baseEnv: {
|
||||
...process.env,
|
||||
PATH: `${binDir}:${process.env.PATH ?? ''}`,
|
||||
MOSAIC_LEASE_BROKER_SOCKET: socket,
|
||||
MOSAIC_RUNTIME_GENERATION: '1',
|
||||
},
|
||||
proxyGate: () =>
|
||||
Promise.resolve({
|
||||
ok: true,
|
||||
report: {
|
||||
binaryPresent: true,
|
||||
binaryPath: '/test/claude-code-proxy',
|
||||
auth: { state: 'valid' },
|
||||
live: true,
|
||||
listenerVerdict: 'ok',
|
||||
needsReauth: false,
|
||||
ok: true,
|
||||
problems: [],
|
||||
},
|
||||
problems: [],
|
||||
}),
|
||||
resolveConfigDir: () => configDir,
|
||||
log: () => {},
|
||||
errorLog: () => {},
|
||||
fail: ((code: number) => {
|
||||
throw new Error(`exit ${code}`);
|
||||
}) as (code: number) => never,
|
||||
});
|
||||
|
||||
expect(execution).toBeDefined();
|
||||
expect(execution!.status, String(execution!.stderr)).toBe(0);
|
||||
expect(JSON.parse(String(execution!.stdout))).toEqual({
|
||||
session_id: expect.stringMatching(/^[a-f0-9]{64}$/),
|
||||
hook_present: true,
|
||||
denied: true,
|
||||
is_yolo: yolo,
|
||||
});
|
||||
},
|
||||
);
|
||||
|
||||
test('Claude and Pi runtime adapters consult the broker for every tool class', async () => {
|
||||
const { socket } = await startBroker();
|
||||
const sessionId = await register(socket);
|
||||
|
||||
expect(runRuntimeGate(socket, sessionId, 'claude', 'Read').status).toBe(0);
|
||||
expect(runRuntimeGate(socket, sessionId, 'claude', 'Bash').status).toBe(2);
|
||||
expect(runRuntimeGate(socket, sessionId, 'pi', 'unknown_custom_tool').status).toBe(2);
|
||||
|
||||
const pending = await beginVerification(socket, sessionId, 'claude');
|
||||
await promote(socket, sessionId, pending.promotion_token!);
|
||||
expect(runRuntimeGate(socket, sessionId, 'claude', 'Bash').status).toBe(0);
|
||||
|
||||
const settings = JSON.parse(await readFile(claudeSettingsPath, 'utf8')) as {
|
||||
hooks: { PreToolUse: Array<{ matcher?: string; hooks: Array<{ command: string }> }> };
|
||||
};
|
||||
expect(
|
||||
settings.hooks.PreToolUse.some(
|
||||
(entry) =>
|
||||
entry.matcher === '.*' &&
|
||||
entry.hooks.some((hook) => hook.command.includes('mutator-gate.py')),
|
||||
),
|
||||
).toBe(true);
|
||||
|
||||
const piExtension = await readFile(piExtensionPath, 'utf8');
|
||||
expect(piExtension).toContain("pi.on('tool_call'");
|
||||
expect(piExtension).toContain('mutator-gate.py');
|
||||
});
|
||||
});
|
||||
Reference in New Issue
Block a user