Files
paperclip/packages/adapters/codex-local/src/server/test.ts
T
Devin Foley b24c6909e8 Harden remote sandbox runtime probes, timeouts, and installs (#5685)
## Thinking Path

> - Paperclip orchestrates AI agents for zero-human companies
> - Each agent runs inside a sandbox environment so its CLI is isolated
from the host
> - Sandbox-backed adapter runs go through a small set of shared helpers
— `ensureAdapterExecutionTargetCommandResolvable`, the sandbox callback
bridge runner, and per-adapter `SANDBOX_INSTALL_COMMAND` strings
> - When standing up new sandbox provider plugins, the existing helpers
timed out, missed install fallbacks, or leaned on assumptions that only
held for E2B
> - Local adapters (`claude-local`, `codex-local`, `gemini-local`,
`opencode-local`) needed slightly hardened probes so they could install
themselves and validate inside *any* remote sandbox transport, not just
E2B
> - This pull request bundles those runtime fixes so future sandbox
provider plugins inherit a working baseline
> - The benefit is that adding a new sandbox provider plugin no longer
requires touching adapter-utils or each local-adapter probe — the
supporting infra is already correct

## What Changed

- `packages/adapter-utils/src/execution-target.ts`: introduce
`DEFAULT_REMOTE_SANDBOX_ADAPTER_TIMEOUT_SEC = 1800` and
`resolveAdapterExecutionTargetTimeoutSec(...)`. Local and SSH adapters
keep the historical "0 means no adapter timeout" behavior;
sandbox-backed runs without an explicit `timeoutSec` get an explicit
30-minute default so remote installs and warm-up don't time out at the
per-RPC default. Plumbed `timeoutSec` through
`ensureAdapterExecutionTargetCommandResolvable` so install probes inside
a sandbox honor adapter-level overrides instead of the bridge's 5-minute
default.
- `packages/adapters/opencode-local/src/index.ts`: switch
`SANDBOX_INSTALL_COMMAND` from `npm install -g opencode-ai` to `curl
-fsSL https://opencode.ai/install | bash`. The npm package reifies four
large prebuilt-binary subpackages in parallel even though only one
matches the host arch; on bandwidth-constrained sandboxes that blew
through the 240s install budget. The official installer fetches one
arch-specific binary and adds `$HOME/.opencode/bin` to PATH via
`~/.bashrc`, which the sandbox-callback-bridge login-shell script
already sources.
- `packages/adapters/{claude,codex,gemini,opencode}-local/`: harden
remote-target probes — pass `--skip-git-repo-check` for Codex when
probing outside a repo, normalize permission flags for Claude, and add
`*.remote.test.ts` coverage that exercises the remote-sandbox path
explicitly for each adapter.
- `packages/adapter-utils/src/sandbox-install-command.{ts,test.ts}`
(new): add `buildSandboxNpmInstallCommand` helper.
`server/src/adapters/registry.ts` + new
`server/src/__tests__/adapter-registry.test.ts`: wire adapter install
commands so they fall back to a writable `$HOME/.local` prefix when
global install isn't available.
- `server/src/__tests__/plugin-worker-manager.test.ts` + new
`server/src/__tests__/fixtures/plugin-worker-delayed.cjs`: pin per-call
timeout overrides so plugin worker exec calls honor the caller's timeout
instead of the worker's default.

## Verification

- `pnpm typecheck`
- `pnpm exec vitest run --no-coverage
packages/adapter-utils/src/execution-target-sandbox.test.ts
packages/adapter-utils/src/sandbox-install-command.test.ts`
- `pnpm exec vitest run --no-coverage
server/src/__tests__/plugin-worker-manager.test.ts
server/src/__tests__/adapter-registry.test.ts
server/src/__tests__/claude-local-adapter-environment.test.ts
server/src/__tests__/claude-local-execute.test.ts
server/src/__tests__/gemini-local-adapter-environment.test.ts`
- `pnpm exec vitest run --no-coverage
packages/adapters/codex-local/src/server/test.remote.test.ts
packages/adapters/opencode-local/src/server/test.remote.test.ts
packages/adapters/codex-local/src/server/codex-args.test.ts
packages/adapters/codex-local/src/server/execute.remote.test.ts
packages/adapters/gemini-local/src/server/execute.remote.test.ts`

All passing locally.

## Risks

- Touches shared `adapter-utils` and several `*-local` adapters. The
30-minute default applies only when both (a) the target is
`remote+sandbox` and (b) no `timeoutSec` is configured — local + SSH
paths are unchanged. New test coverage was added alongside each behavior
change to pin the contracts.
- Switching OpenCode's install command to the official installer is a
behavior change for any operator running OpenCode inside a remote
sandbox. Local installs are unaffected (the `SANDBOX_INSTALL_COMMAND`
only runs when an adapter is being installed inside a sandbox).
- Low risk overall — no migrations, no API surface change.

## Model Used

- Provider: Anthropic
- Model: Claude Opus 4.7 (1M context)
- Capabilities used: extended reasoning, tool use (Read/Edit/Bash/Grep),
no code execution beyond local repo commands

## Checklist

- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [ ] If this change affects the UI, I have included before/after
screenshots — N/A, no UI change
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [x] I will address all Greptile and reviewer comments before
requesting merge

Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-05-11 00:31:54 -07:00

392 lines
14 KiB
TypeScript

import type {
AdapterEnvironmentCheck,
AdapterEnvironmentTestContext,
AdapterEnvironmentTestResult,
} from "@paperclipai/adapter-utils";
import {
asString,
parseObject,
ensurePathInEnv,
} from "@paperclipai/adapter-utils/server-utils";
import {
ensureAdapterExecutionTargetCommandResolvable,
ensureAdapterExecutionTargetDirectory,
maybeRunSandboxInstallCommand,
runAdapterExecutionTargetProcess,
describeAdapterExecutionTarget,
resolveAdapterExecutionTargetCwd,
prepareAdapterExecutionTargetRuntime,
} from "@paperclipai/adapter-utils/execution-target";
import fs from "node:fs/promises";
import path from "node:path";
import os from "node:os";
import { parseCodexJsonl } from "./parse.js";
import { SANDBOX_INSTALL_COMMAND } from "../index.js";
import { codexHomeDir, readCodexAuthInfo } from "./quota.js";
import { buildCodexExecArgs } from "./codex-args.js";
import { prepareManagedCodexHome } from "./codex-home.js";
function summarizeStatus(checks: AdapterEnvironmentCheck[]): AdapterEnvironmentTestResult["status"] {
if (checks.some((check) => check.level === "error")) return "fail";
if (checks.some((check) => check.level === "warn")) return "warn";
return "pass";
}
function isNonEmpty(value: unknown): value is string {
return typeof value === "string" && value.trim().length > 0;
}
function firstNonEmptyLine(text: string): string {
return (
text
.split(/\r?\n/)
.map((line) => line.trim())
.find(Boolean) ?? ""
);
}
function commandLooksLike(command: string, expected: string): boolean {
const base = path.basename(command).toLowerCase();
return base === expected || base === `${expected}.cmd` || base === `${expected}.exe`;
}
function summarizeProbeDetail(stdout: string, stderr: string, parsedError: string | null): string | null {
const raw = parsedError?.trim() || firstNonEmptyLine(stderr) || firstNonEmptyLine(stdout);
if (!raw) return null;
const clean = raw.replace(/\s+/g, " ").trim();
const max = 240;
return clean.length > max ? `${clean.slice(0, max - 1)}` : clean;
}
const CODEX_AUTH_REQUIRED_RE =
/(?:not\s+logged\s+in|login\s+required|authentication\s+required|unauthorized|invalid(?:\s+or\s+missing)?\s+api(?:[_\s-]?key)?|openai[_\s-]?api[_\s-]?key|api[_\s-]?key.*required|please\s+run\s+`?codex\s+login`?)/i;
async function prepareCodexHelloProbe(input: {
runId: string;
companyId: string;
target: AdapterEnvironmentTestContext["executionTarget"] | null;
targetIsRemote: boolean;
cwd: string;
command: string;
args: string[];
env: Record<string, string>;
probeApiKey: string | null;
}): Promise<{
command: string;
args: string[];
env: Record<string, string>;
cleanup: () => Promise<void>;
}> {
let preparedRuntime: Awaited<ReturnType<typeof prepareAdapterExecutionTargetRuntime>> | null = null;
let preparedRuntimeWorkspaceLocalDir: string | null = null;
const cleanup = async () => {
await preparedRuntime?.restoreWorkspace().catch(() => {});
if (preparedRuntimeWorkspaceLocalDir) {
await fs.rm(preparedRuntimeWorkspaceLocalDir, { recursive: true, force: true }).catch(() => {});
}
};
if (input.targetIsRemote && !input.probeApiKey) {
const managedHome = await prepareManagedCodexHome(process.env, async () => {}, input.companyId, {
apiKey: null,
});
preparedRuntimeWorkspaceLocalDir = await fs.mkdtemp(
path.join(os.tmpdir(), `paperclip-codex-envtest-${input.runId}-`),
);
preparedRuntime = await prepareAdapterExecutionTargetRuntime({
runId: input.runId,
target: input.target,
adapterKey: "codex",
workspaceLocalDir: preparedRuntimeWorkspaceLocalDir,
// Pass `input.cwd` as the base (not a pre-built per-run subdir).
// `prepareRemoteManagedRuntime` itself appends
// `.paperclip-runtime/runs/<runId>/workspace` to whatever it gets, so
// pre-building a per-run path here would double-nest the run ID.
workspaceRemoteDir: input.cwd,
installCommand: SANDBOX_INSTALL_COMMAND,
detectCommand: input.command,
assets: [
{
key: "home",
localDir: managedHome,
followSymlinks: true,
},
],
});
return {
command: input.command,
args: input.args,
env: preparedRuntime.assetDirs.home
? { ...input.env, CODEX_HOME: preparedRuntime.assetDirs.home }
: { ...input.env },
cleanup,
};
}
if (input.probeApiKey) {
const probeHome = input.targetIsRemote
? path.posix.join(input.cwd, ".paperclip-runtime", "codex", `probe-home-${input.runId}`)
: path.join(os.tmpdir(), `paperclip-codex-probe-${input.runId}`);
return {
command: "sh",
args: [
"-c",
'set -e; mkdir -p "$CODEX_HOME"; umask 077; printf "%s" "$_PAPERCLIP_CODEX_AUTH_JSON" > "$CODEX_HOME/auth.json"; unset _PAPERCLIP_CODEX_AUTH_JSON; trap \'rm -rf "$CODEX_HOME"\' EXIT INT TERM; "$0" "$@"',
input.command,
...input.args,
],
env: {
...input.env,
CODEX_HOME: probeHome,
_PAPERCLIP_CODEX_AUTH_JSON: JSON.stringify({ OPENAI_API_KEY: input.probeApiKey }),
},
cleanup,
};
}
return {
command: input.command,
args: input.args,
env: { ...input.env },
cleanup,
};
}
export async function testEnvironment(
ctx: AdapterEnvironmentTestContext,
): Promise<AdapterEnvironmentTestResult> {
const checks: AdapterEnvironmentCheck[] = [];
const config = parseObject(ctx.config);
const command = asString(config.command, "codex");
const target = ctx.executionTarget ?? null;
const targetIsRemote = target?.kind === "remote";
const targetIsSandbox = target?.kind === "remote" && target.transport === "sandbox";
const cwd = resolveAdapterExecutionTargetCwd(target, asString(config.cwd, ""), process.cwd());
const targetLabel = targetIsRemote
? ctx.environmentName ?? describeAdapterExecutionTarget(target)
: null;
const runId = `codex-envtest-${Date.now()}-${Math.random().toString(16).slice(2)}`;
if (targetLabel) {
checks.push({
code: "codex_environment_target",
level: "info",
message: `Probing inside environment: ${targetLabel}`,
});
}
try {
await ensureAdapterExecutionTargetDirectory(runId, target, cwd, {
cwd,
env: {},
createIfMissing: true,
});
checks.push({
code: "codex_cwd_valid",
level: "info",
message: `Working directory is valid: ${cwd}`,
});
} catch (err) {
checks.push({
code: "codex_cwd_invalid",
level: "error",
message: err instanceof Error ? err.message : "Invalid working directory",
detail: cwd,
});
}
const envConfig = parseObject(config.env);
const env: Record<string, string> = {};
for (const [key, value] of Object.entries(envConfig)) {
if (typeof value === "string") env[key] = value;
}
const runtimeEnv = ensurePathInEnv({ ...process.env, ...env });
const installCheck = await maybeRunSandboxInstallCommand({
runId,
target,
adapterKey: "codex",
installCommand: SANDBOX_INSTALL_COMMAND,
detectCommand: command,
env,
});
if (installCheck) checks.push(installCheck);
try {
await ensureAdapterExecutionTargetCommandResolvable(command, target, cwd, runtimeEnv);
checks.push({
code: "codex_command_resolvable",
level: "info",
message: `Command is executable: ${command}`,
});
} catch (err) {
checks.push({
code: "codex_command_unresolvable",
level: "error",
message: err instanceof Error ? err.message : "Command is not executable",
detail: command,
});
}
const configOpenAiKey = env.OPENAI_API_KEY;
const hostOpenAiKey = targetIsRemote ? undefined : process.env.OPENAI_API_KEY;
if (isNonEmpty(configOpenAiKey) || isNonEmpty(hostOpenAiKey)) {
const source = isNonEmpty(configOpenAiKey) ? "adapter config env" : "server environment";
checks.push({
code: "codex_openai_api_key_present",
level: "info",
message: "OPENAI_API_KEY is set for Codex authentication.",
detail: `Detected in ${source}.`,
});
} else if (!targetIsRemote) {
// Local-only auth file check. On remote targets, the probe will surface
// any missing-auth errors directly from the remote `codex` invocation.
const codexHome = isNonEmpty(env.CODEX_HOME) ? env.CODEX_HOME : undefined;
const codexAuth = await readCodexAuthInfo(codexHome).catch(() => null);
if (codexAuth) {
checks.push({
code: "codex_native_auth_present",
level: "info",
message: "Codex is authenticated via its own auth configuration.",
detail: codexAuth.email ? `Logged in as ${codexAuth.email}.` : `Credentials found in ${path.join(codexHome ?? codexHomeDir(), "auth.json")}.`,
});
} else {
checks.push({
code: "codex_openai_api_key_missing",
level: "warn",
message: "OPENAI_API_KEY is not set. Codex runs may fail until authentication is configured.",
hint: "Set OPENAI_API_KEY in adapter env, shell environment, or run `codex auth` to log in.",
});
}
}
const canRunProbe =
checks.every((check) => check.code !== "codex_cwd_invalid" && check.code !== "codex_command_unresolvable");
if (canRunProbe) {
if (!commandLooksLike(command, "codex")) {
checks.push({
code: "codex_hello_probe_skipped_custom_command",
level: "info",
message: "Skipped hello probe because command is not `codex`.",
detail: command,
hint: "Use the `codex` CLI command to run the automatic login and installation probe.",
});
} else {
const execArgs = buildCodexExecArgs(
{ ...config, fastMode: false },
{ skipGitRepoCheck: targetIsSandbox },
);
const args = execArgs.args;
if (execArgs.fastModeIgnoredReason) {
checks.push({
code: "codex_fast_mode_unsupported_model",
level: "warn",
message: execArgs.fastModeIgnoredReason,
hint: "Switch the agent model to GPT-5.4 or enter a manual model ID to enable Codex Fast mode.",
});
}
if (targetIsSandbox) {
checks.push({
code: "codex_git_repo_check_skipped",
level: "info",
message: "Added --skip-git-repo-check for sandbox hello probes.",
hint: "Codex requires an explicit trust bypass in headless remote sandbox workspaces.",
});
}
// Codex CLI (>= 0.122) ignores the OPENAI_API_KEY env var and only reads
// credentials from $CODEX_HOME/auth.json. When we have a key available,
// wrap the probe with a shell that materializes a per-run auth.json so
// the CLI can authenticate. The key content is passed via env (not on
// the command line) to avoid leaking it into process listings.
const probeApiKey = isNonEmpty(configOpenAiKey)
? configOpenAiKey
: isNonEmpty(hostOpenAiKey)
? hostOpenAiKey
: null;
const preparedProbe = await prepareCodexHelloProbe({
runId,
companyId: ctx.companyId,
target,
targetIsRemote,
cwd,
command,
args,
env,
probeApiKey,
});
try {
const probe = await runAdapterExecutionTargetProcess(
runId,
target,
preparedProbe.command,
preparedProbe.args,
{
cwd,
env: preparedProbe.env,
timeoutSec: 45,
graceSec: 5,
stdin: "Respond with hello.",
onLog: async () => {},
},
);
const parsed = parseCodexJsonl(probe.stdout);
const detail = summarizeProbeDetail(probe.stdout, probe.stderr, parsed.errorMessage);
const authEvidence = `${parsed.errorMessage ?? ""}\n${probe.stdout}\n${probe.stderr}`.trim();
if (probe.timedOut) {
checks.push({
code: "codex_hello_probe_timed_out",
level: "warn",
message: "Codex hello probe timed out.",
hint: "Retry the probe. If this persists, verify Codex can run `Respond with hello` from this directory manually.",
});
} else if ((probe.exitCode ?? 1) === 0) {
const summary = parsed.summary.trim();
const hasHello = /\bhello\b/i.test(summary);
checks.push({
code: hasHello ? "codex_hello_probe_passed" : "codex_hello_probe_unexpected_output",
level: hasHello ? "info" : "warn",
message: hasHello
? "Codex hello probe succeeded."
: "Codex probe ran but did not return `hello` as expected.",
...(summary ? { detail: summary.replace(/\s+/g, " ").trim().slice(0, 240) } : {}),
...(hasHello
? {}
: {
hint: "Try the probe manually (`codex exec --json -` then prompt: Respond with hello) to inspect full output.",
}),
});
} else if (CODEX_AUTH_REQUIRED_RE.test(authEvidence)) {
checks.push({
code: "codex_hello_probe_auth_required",
level: "warn",
message: "Codex CLI is installed, but authentication is not ready.",
...(detail ? { detail } : {}),
hint: probeApiKey
? "OPENAI_API_KEY was provided but Codex still rejected the request. Verify the key is valid for the OpenAI Responses API (e.g. `curl -H \"Authorization: Bearer $OPENAI_API_KEY\" https://api.openai.com/v1/models`), or run `codex login` and seed `~/.codex/auth.json`."
: "Codex CLI does not read OPENAI_API_KEY from the environment; set OPENAI_API_KEY in this adapter's config (so Paperclip writes it to `$CODEX_HOME/auth.json`) or run `codex login` on the host first.",
});
} else {
checks.push({
code: "codex_hello_probe_failed",
level: "error",
message: "Codex hello probe failed.",
...(detail ? { detail } : {}),
hint: "Run `codex exec --json -` manually in this working directory and prompt `Respond with hello` to debug.",
});
}
} finally {
await preparedProbe.cleanup();
}
}
}
return {
adapterType: ctx.adapterType,
status: summarizeStatus(checks),
checks,
testedAt: new Date().toISOString(),
};
}