@cachly-dev/mcp-server@0.10.185
Solve it once. From then on, your AI knows. Experience for coding agents, EU-hosted: 124 MCP tools for Claude Code, Cursor, Copilot & Windsurf. Learns from every bug fix, commit and CI run.
AI Security Review
scanned 10d ago · by lpm-firewall-aiThis is a first-party agent-hook setup capability, activated through the setup command, that can affect agent context and exchange data during later sessions. The source does not establish a malicious transfer, so warn rather than block.
Decision evidence
public snapshot- The manifest defines a postinstall hook, and its script prints an instruction to run the setup command.
- The setup branch requires an explicit setup or autopilot command and can register Claude hooks.
- Setup adds package-owned hooks to Claude settings for sessions, prompts, file operations, and stop events.
- The hooks retrieve Cachly memories and send learned data to the configured Cachly API.
- The inspected credential and agent references concern Cachly credentials, Cachly plugin setup, team instructions, and local git configuration.
- The postinstall hook displays the setup command; the inspected source performs editor-hook setup during that CLI flow.
- The hook requests use the Cachly API, defaulting to api.cachly.dev, with a Cachly JWT; these excerpts do not show transfer to an unrelated recipient.
Source & flagged code
12 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgFindings
Affected versions and remediation
This report applies to @cachly-dev/mcp-server@0.10.185.
See all versions of @cachly-dev/mcp-server for other recorded verdicts.
- Review the evidence and your use of @cachly-dev/mcp-server@0.10.185 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.