All @mastra/voice-sarvam versions

@mastra/[email protected] rejected Risk: 100 Apache-2.0

@mastra/voice-sarvam @1.0.2

rejected
This version was rejected. It did not pass GreenFlagged's security review and is not served by the registry. The findings and risk dispositions below explain why.
npm Repository Homepage Tarball
100
Risk Score
Apache-2.0
License
No
Install Scripts
1
Dependencies
11
Dev Dependencies
182.5 KB
Package Size
Published

Mastra Sarvam AI voice integration

Maintainers

ehindero

Keywords

mastrasarvamttssttindian-languagesspeech-to-texttext-to-speechspeech-recognition

Dependencies (1)

PackageConstraintRegistry Status
easy-day-js ^1.11.21 auto_approved

Dev Dependencies (11)

PackageConstraintRegistry Status
zod ^4.4.3 auto_approved
tsup ^8.5.1 auto_approved
eslint ^10.2.1 auto_approved
vitest 4.1.5 No greenflagged match
@vitest/ui 4.1.5 auto_approved
typescript ^6.0.3 auto_approved
@types/node 22.19.15 auto_approved
@internal/lint 0.0.100 Not imported
@internal/voice 0.0.1 Not imported
@vitest/coverage-v8 4.1.5 auto_approved
@internal/types-builder 0.0.75 Not imported

Transitive Dependency Tree

1 transitive deps max depth 1
  ├─ easy-day-js ^1.11.21 → 1.11.21

Changes from v1.0.1

Dependency Changes

ChangePackageVersion
added easy-day-js ^1.11.21

File Changes

0 added 0 removed 1 modified size delta: +.1 KB

Risk Dispositions (2 applicable to this version, 0 other)

Accepted rules are downgraded to INFO on future analyses; rejected rules escalate to CRITICAL.

Rule Source Disposition Author Reason
regressed-provenance provenance reject AI AI (provenance): Prior versions had CI/CD provenance; manual publish without attestation is a strong compromise indicator for this package.
unvetted-dep:easy-day-js dependencies reject AI AI (dependencies): Newly added unvetted dep with no legitimate use (also phantom); matches axios attack vector.

SAST Findings (3)

HIGH Provenance attestation missing — previous versions had it provenance

This version was published without provenance, but prior versions were published via CI/CD with attestations. This is a strong signal of a potential account compromise or unauthorized publish. The axios attack (March 2026) exhibited exactly this pattern.

HIGH New obfuscated file: dist/_types/@internal_voice/dist/_types/@internal_ai-sdk-v5/dist/index.d.ts source-diff

Newly added source file contains lines over 3000 chars, suggesting minified or obfuscated code. New obfuscated files are a strong attack indicator.

INFO Publisher changed: GitHub Actions → ehindero (on 2026-06-17, known maintainer) provenance

This version was published by a different npm account (ehindero) than the most recent previously approved version (GitHub Actions) on 2026-06-17, but ehindero 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.

Review Summary

Risk score: 100 (capped from 103). Findings: 2 high (+50), 5 medium (+50), 1 low (+3), 2 info (+0).

Published to npm: