@papi-ai/server@0.7.72
PAPI MCP server — AI-powered sprint planning, build execution, and strategy review for software projects
AI Security Review
scanned 3mo ago · by lpm-firewall-aiThe static malicious finding overstates package-aligned Git and telemetry behavior; command output is not sent to a third party. Policy still calls for a warning for explicit user-command AI-agent configuration mutation.
Decision evidence
public snapshot- dist/index.js: CLI setup writes a PAPI MCP entry with API credentials to .mcp.json.
- dist/index.js: init tool can write PAPI MCP configuration for several AI clients, including user-scoped Codex/Hermes configs.
- dist/index.js: runtime creates a persistent ~/.papi/install-id.json and sends telemetry when enabled.
- dist/index.js: release/workflow functions invoke git commands and can commit or push project changes.
- package.json has no preinstall, install, or postinstall lifecycle hook.
- Configuration mutation is exposed through explicit setup/init commands or MCP tools, not package installation.
- Telemetry payload construction contains tool/milestone metadata and project identifiers; no command output or harvested file content is sent.
- Network hosts are package-aligned PAPI/Supabase services and npm registry.
- No eval, arbitrary shell string execution, payload download/execution, or credential harvesting found.
Source & flagged code
8 flaggedPackage source references child process execution.
dist/backfill-cycle-metrics.jsView on unpkg · L78A single source file combines environment access, network access, and code or shell execution; review context before blocking.
dist/backfill-cycle-metrics.jsView on unpkg · L4268Findings
Affected versions and remediation
This report applies to @papi-ai/server@0.7.72.
See all versions of @papi-ai/server for other recorded verdicts.
- Review the evidence and your use of @papi-ai/server@0.7.72 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.