registry / @bfrs/agentic-components / 0.8.3
@bfrs/agentic-components@0.8.3
Shiprocket agentic component library and documentation showcase.
AI Security Review
scanned 6d ago · by lpm-firewall-aiThe postinstall hook makes unconfirmed changes to consumer agent instructions and build scripts, so residual agent integration risk remains. Its observed changes are package-specific context and lint setup, with no concrete attack behavior identified.
Static reason
High-risk behavior combination matched malicious policy.
Trigger
Installing the package in a consumer project without setting its setup opt-out variable.
Impact
Consumer agent instructions and build behavior are changed without a separate setup confirmation, creating residual agent integration risk; no concrete malicious effect was identified.
Mechanism
The hook copies the package context document, adds a reference to it in AGENTS.md or CLAUDE.md, and adds the package lint command to the consumer package scripts and existing build scripts.
AI rationale
The postinstall hook makes unconfirmed changes to consumer agent instructions and build scripts, so residual agent integration risk remains. Its observed changes are package-specific context and lint setup, with no concrete attack behavior identified.
Evidence
package.jsonscripts/postinstall.cjsBFRS_AGENTIC_COMPONENTS.mdAGENTS.mdCLAUDE.md
Decision evidence
public snapshotAI called this Suspicious at 91.0% confidence as Dangerous Capability with medium false-positive risk.
Evidence for warning
- package.json runs scripts/postinstall.cjs automatically during installation.
- The hook targets consumer installs and supports an environment-variable opt-out; absent that opt-out, it proceeds with setup.
- The hook appends a reference to its package-owned context file in consumer agent instruction files, including AGENTS.md and CLAUDE.md.
- The hook copies that context file into the consumer project and adds its lint command to the consumer package scripts, including existing build scripts.
- The package describes the context and lint guard as setup for using its component library.
Evidence against
- The inspected hook shows bounded documentation and lint-script changes; no credential collection, network transfer, or remote code execution was identified.
- The setup is package-owned and tied to the library’s linting purpose, which limits the evidence for malicious intent.
Behavioral surface
ChildProcessEnvironmentVarsFilesystem
Source & flagged code
6 flagged · loading sourcepackage.jsonView file
•scripts.postinstall = node scripts/postinstall.cjs
High
Install Time Lifecycle Scripts
Package defines install-time lifecycle scripts.
package.jsonView on unpkg•scripts.postinstall = node scripts/postinstall.cjs
Medium
Ambiguous Install Lifecycle Script
Install-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkg54"scripts": {
"postinstall": "node scripts/postinstall.cjs",
"dev": "vite --host 0.0.0.0",
"build": "npm ru
Medium
Suspicious Lifecycle Evidence
package.json runs scripts/postinstall.cjs automatically during installation.
package.jsonView on unpkg · L54scripts/postinstall.cjsView file
Findings
1 Critical2 High5 Medium4 Low
CriticalAi Agent Control Hijackscripts/postinstall.cjs
HighInstall Time Lifecycle Scriptspackage.json
HighSemantic Analysis Limiteddist/fonts.css
MediumAmbiguous Install Lifecycle Scriptpackage.json
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
MediumSuspicious Lifecycle Evidencepackage.json
MediumAi Review Evidencescripts/postinstall.cjs
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to @bfrs/agentic-components@0.8.3.
See all versions of @bfrs/agentic-components for other recorded verdicts.
- Review the evidence and your use of @bfrs/agentic-components@0.8.3 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.