registry / @readwise/cli / 0.5.8

@readwise/cli@0.5.8

Command-line interface for Readwise and Reader

AI Security Review

scanned 9h ago · by lpm-firewall-ai

The interactive TUI can execute shell syntax embedded in an MCP-supplied card URL after the user selects Open. Separately, explicit skills installation writes remotely fetched skill files into selected agent directories.

Static reason

No blocking static signals were detected.

Trigger

Authenticated user opens a result card in the TUI; agent writes require explicit readwise skills install.

Impact

A malicious URL returned by the MCP service could execute commands as the CLI user.

Mechanism

Shell command injection through exec() of an unescaped remote URL

Rationale

Source contains a concrete remote-data-to-shell execution path, but no malicious lifecycle behavior or covert collection/exfiltration. The package should be warned for remediation rather than blocked as malware.

Evidence

package.jsonsrc/index.tssrc/auth.tssrc/mcp.tssrc/commands.tssrc/config.tssrc/skills.tssrc/tui/app.tsREADME.md

Network endpoints4

mcp2.readwise.io/mcpreadwise.io/o/.well-known/oauth-authorization-serverapi.github.com/repos/readwiseio/readwise-skillsraw.githubusercontent.com/readwiseio/readwise-skills/master

@readwise/cli@0.5.8

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Source & flagged code