@picovoice/picollm-node
Picovoice picoLLM Node.js binding
Supply chain provenance
Status for the latest visible version.
Without SLSA provenance there is no cryptographic link between this tarball and the public source — the axios compromise (March 2026) relied on exactly this gap.
Maintainers
Keywords
Accepted risks
Findings the reviewer chose to accept rather than block on.
| Source | Rule | Reason | Accepted by | When |
|---|---|---|---|---|
| npm-metadata | bundled-binaries | AI (npm-metadata): Picovoice ships prebuilt platform-specific native .node addons for all their SDKs; bundled binaries are the documented distribution mechanism for this native binding. | ai | |
| semgrep | semgrep:dynamic-require | AI (semgrep): Dynamic require(libraryPath) is the standard pattern for loading platform-specific native .node addons; stable and expected for this package. | ai |
Versions (showing 4 of 4)
| Version | Deps | Published |
|---|---|---|
| 2.1.2 | 0 / 14 | |
| 2.1.0 | 0 / 14 | |
| 2.0.1 | 0 / 13 | |
| 2.0.0 | 0 / 13 |
v2.1.2
2 findingsPackage was published without Sigstore provenance. Consider requesting the maintainer enable provenance via CI/CD.
This version was published by a different npm account (matt200-ok) than the most recent previously approved version (ilavery) on 2026-06-09, but matt200-ok is listed as a maintainer on prior approved versions (matched on name). This looks like a manual publish by a known maintainer rather than a publisher change. Recorded as INFO for audit trail.
v2.1.0
2 findingsPackage contains compiled binaries that could be backdoors: • lib/windows/amd64/pv_ypu_impl_cuda_picollm.dll • lib/linux/x86_64/pv_picollm.node • lib/mac/arm64/pv_picollm.node • lib/mac/x86_64/pv_picollm.node • lib/raspberry-pi/cortex-a53-aarch64/pv_picollm.node • lib/raspberry-pi/cortex-a53/pv_picollm.node • lib/raspberry-pi/cortex-a72-aarch64/pv_picollm.node • lib/raspberry-pi/cortex-a72/pv_picollm.node • lib/raspberry-pi/cortex-a76-aarch64/pv_picollm.node • lib/raspberry-pi/cortex-a76/pv_picollm.node ... and 2 more
Package was published without Sigstore provenance. Only ~12% of npm packages have provenance, so this is common but not ideal.
v2.0.1
1 findingPackage was published without Sigstore provenance. Only ~12% of npm packages have provenance, so this is common but not ideal.
v2.0.0
1 findingPackage was published without Sigstore provenance. Only ~12% of npm packages have provenance, so this is common but not ideal.