Your RSA-2048 keys break in 2030. Find every one of them before attackers do.
Part of a larger attack: @onereach published 78 malicious packages. See the full campaign →
Malicious package

@onereach/step-voicenpm

@onereach/step-voice is a confirmed malicious npm package (MAL-2026-11655) that steals credentials and exfiltrates sensitive data (malicious versions 7.0.32, 7.0.33, 7.0.34). Do not install it — remove it immediately and rotate any exposed credentials.

Malicious code in @onereach/step-voice (npm)

MAL-2026-11655
Immediate action
Remove the package, then rotate any secrets the build/runtime could reach.
npm uninstall @onereach/step-voice

What this malware does

npm/@onereach/step-voice is affected by the large-scale, self-propagating npm supply-chain worm of 2026-08-04 (the "Shai-Hulud: Here We Go Again" wave) — the same campaign that began with the compromise of the keyv and cacheable maintainer account. The listed version(s) were trojanized and republished by the worm after it reached an npm publish token belonging to a maintainer in this namespace; the payload enumerates every package a stolen token controls and republishes each with the same hook, so many packages under this scope were poisoned in the same short window. Every poisoned release adds a preinstall hook ("preinstall": "node setup.mjs") that runs on a bare npm install, before any project code. setup.mjs downloads a standalone Bun runtime and runs a byte-identical, heavily obfuscated ~728 KB second-stage credential stealer (shipped as Math_Symbol.js / math_init.js). It harvests GitHub, npm, AWS, GCP, Azure, HashiCorp Vault and Kubernetes credentials plus generic secrets and private keys (TruffleHog-style sweep), reads CI/CD secrets and identifies build runners, then republishes further packages the stolen token can reach. Rather than a fixed command-and-control host, it exfiltrates stolen findings to attacker-created GitHub repositories (descriptions reading "Shai-Hulud: Here We Go Again") and over DNS. Treat any environment that installed an affected version (with install scripts enabled) as compromised: rotate and revoke all reachable credentials (npm and GitHub tokens, cloud keys, Vault/Kubernetes secrets, and CI org/repo secrets). Part of the August 2026 npm worm that poisoned 400+ packages across many organizations.

@onereach/step-voice 7.0.32 ships a preinstall hook (node setup.mjs) that, on npm install, downloads the Bun runtime from oven-sh's GitHub release for v1.3.13, then uses that Bun binary to execute a sibling file math_init.js (727 KB) via execFileSync(bunPath, [math_init.js]). setup.mjs itself is obfuscated with _0x... hex-suffixed identifiers and dictionary-indirection to hide its network and process calls. math_init.js is a Bun-compiled bundle (// @bun @bun-cjs header) with a shuffled numeric constant array and an index-rotation decoder — an opaque payload that is not human-inspectable as Node source and whose filename is unrelated to the package's declared purpose (Onereach.ai voice-conversation steps). Integrity signals compound the concern: package.json declares files: ["/dst"] (only the dst/ directory should be published) yet the tarball root contains both setup.mjs and math_init.js; the declared main: index.js does not exist in the tarball. The delivery shape — install-time fetch of an alternate runtime used to execute an adjacent obfuscated compiled blob unrelated to the package's purpose — is the alternate-runtime-dropper pattern.

Any computer that has this package installed or running should be considered fully compromised. All secrets and keys stored on that computer should be rotated immediately from a different computer. The package should be removed, but as full control of the computer may have been given to an outside entity, there is no guarantee that removing the package will remove all malicious software resulting from installing it.

Malicious versions

3 flagged
7.0.327.0.337.0.34

Indicators of compromise (SHA-256)

08e4dae9b5627ad354ac1cc3c6e0e92ec321bb2f8b730fbe8df5e65ce22767df
52df6e500569e997f3e4b3e3e590d7abd4566c43ea67b70258e13261bc8f4e87
0e9aa53b0e836ed0fc826094a55a1016b42644dc1dfc05a4191d2e236f2a2a7c
44111716a3e4a02f9ac39d153dbcd05d911c0c39a835ac2b6592686df232ec5a

Detection & response playbook

Credential / info stealer
  1. Find it

    Scan your lockfiles (package-lock.json, pnpm-lock.yaml, yarn.lock, requirements.txt, poetry.lock, etc.) and build artifacts for @onereach/step-voice (3 malicious versions). O3 Security's supply-chain scanner checks every dependency against known-malicious package intelligence at install time and in CI, flagging @onereach/step-voice across your stack and pipelines.

  2. If you installed it — respond

    @onereach/step-voice is built to steal secrets, so assume every credential the build or runtime could read is compromised. Remove it from your project and lockfile, then rotate ALL exposed secrets — npm/registry tokens, cloud keys, CI/CD secrets, SSH keys, and any .env values — from a known-clean machine. Audit logs for unauthorized use of those credentials.

  3. Did it already run?

    If @onereach/step-voice was ever installed, its post-install/runtime payload may have already executed. O3's L7 egress monitoring and runtime eBPF sensors detect the credential exfiltration or command-and-control callback after install and block the malicious outbound channel, so you catch and contain the actual compromise — not just the presence of the package.

  4. How O3 protects you

    O3 blocks @onereach/step-voice before install through its supply-chain scanner, and if it has already run, detects and severs the exfiltration or C2 callback at runtime through L7 egress monitoring and eBPF.

Frequently asked questions

No. @onereach/step-voice on npm has been identified as a malicious package (versions 7.0.32, 7.0.33, 7.0.34 flagged). It should be removed immediately — do not install or keep it in your dependency tree.

Campaign

GHSA-qc9m-43v9-6f42IN-MAL-2026-013814IN-MAL-2026-014028IN-MAL-2026-014111

References

Credits

  • Aikido Security · finder
  • Amazon Inspector · finder
  • SafeDep · finder
  • Socket Threat Research Team · finder

Detect & block this

O3 blocks @onereach/step-voice-class packages before install and in CI — and if it already ran, its runtime egress monitoring catches the credential exfiltration and severs the channel.

Explore

@onereach/step-voice (npm) malicious package — MAL-2026-11655 | O3 Security