@contentful/experience-design-system-cli@2.17.0
Contentful Experiences design system import CLI
AI Security Review
scanned 3mo ago · by lpm-firewall-aiSource inspection found no malicious install hook, stealth persistence, credential exfiltration, or unsolicited control-surface mutation. The unconditional sandbox-bypass flag for an external coding agent remains an unresolved dangerous capability.
Static reason
No blocking static signals were detected.; previous stored version diff introduced dangerous source
Trigger
User runs generate/import with Codex selected, or interactively confirms setup actions.
Impact
A malicious or prompt-injected input could receive the selected agent's unrestricted local workspace access.
Mechanism
Spawns external coding-agent CLI with unrestricted Codex execution flag.
AI rationale
Source inspection found no malicious install hook, stealth persistence, credential exfiltration, or unsolicited control-surface mutation. The unconditional sandbox-bypass flag for an external coding agent remains an unresolved dangerous capability.
Evidence
package.jsonbin/cli.jsdist/src/generate/agent-runner.jsdist/src/generate/command.jsdist/src/setup/command.jsdist/src/apply/api-client.jsdist/src/credentials-store.js
Network endpoints2
api.contentful.comraw.githubusercontent.com/nvm-sh/nvm/v0.40.4/install.sh
Decision evidence
public snapshotAI called this Suspicious at 90.0% confidence as Unknown with low false-positive risk.
Evidence for warning
- dist/src/generate/agent-runner.js runs Codex with --dangerously-bypass-approvals-and-sandbox.
- The same runner spawns configured external agent binaries with generated prompts.
- dist/src/setup/command.js offers confirmed setup actions that install tools and can pipe an nvm URL to bash.
Evidence against
- package.json has no preinstall, install, or postinstall hooks.
- CLI entry imports the application only after a source-only rebuild check.
- Agent execution is reached through explicit generate/import commands after agent selection.
- Contentful API requests use operator-provided credentials and documented API paths.
- PathValidationStep only reads a user-selected project directory to count files.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemNetworkShell
HighEntropyStringsObfuscatedUrlStrings
Source & flagged code
2 flaggeddist/src/import/tui/steps/PathValidationStep.jsView file
•stage = ast_semantic_analysis; reason = ast_alias_growth_limit_exceeded; limitedFiles = 1
High
Semantic Analysis Limited
A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
dist/src/import/tui/steps/PathValidationStep.jsView on unpkgdist/src/generate/agent-runner.jsView file
•matchType = previous_version_dangerous_delta
matchedPackage = @contentful/experience-design-system-cli@2.16.1
matchedIdentity = npm:[redacted]:2.16.1
similarity = 0.933
summary = stored previous version shares package body but lacks this dangerous source file
High
Previous Version Dangerous Delta
This package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/src/generate/agent-runner.jsView on unpkgFindings
2 High3 Medium5 Low
HighSemantic Analysis Limiteddist/src/import/tui/steps/PathValidationStep.js
HighPrevious Version Dangerous Deltadist/src/generate/agent-runner.js
MediumNetwork
MediumEnvironment Vars
MediumStructural Risk Force Deep Review
LowScripts Present
LowFilesystem
LowObfuscated
LowHigh Entropy Strings
LowUrl Strings
Affected versions and remediation
This report applies to @contentful/experience-design-system-cli@2.17.0.
See all versions of @contentful/experience-design-system-cli for other recorded verdicts.
- Review the evidence and your use of @contentful/experience-design-system-cli@2.17.0 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.