@cotal-ai/cli@0.122.0
Cotal mesh CLI: up, join, watch, console, spawn, mint, channels, history, ext.
AI Security Review
scanned 6d ago · by lpm-firewall-aiThe manifest has no install lifecycle scripts, and the reviewed source ties subprocesses, network requests, and agent-skill writes to CLI commands. Runtime telemetry is disclosed, supports an off setting, and the inspected telemetry code describes its data limits.
Decision evidence
public snapshot- The package entry points load the CLI command registry; the manifest declares no install lifecycle scripts.
- Runtime telemetry respects the configured off level and is described as anonymous, with a notice and documented opt-out.
- Telemetry is armed for a command invocation and sent on process exit; the source documents the fields and says argument values, paths, and names are excluded.
- Agent skill files are reconciled by the explicit `cotal setup` command.
- The global npm install is offered only for an npx setup run without `cotal` on PATH, with an interactive confirmation or the documented non-interactive default.
- Feedback submission is a command that builds a report from supplied inputs and sends it to the configured or Cotal intake.
- The CLI uses network access and writes agent skill files during setup, but the inspected paths describe these as runtime command behaviors rather than package install or import-time actions.
- Regular telemetry sends a daily install summary and command event; the code exposes an off setting and documents the collection and opt-out.
Source & flagged code
7 flaggedPackage source references child process execution.
dist/seed/reconcile.jsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @cotal-ai/cli@0.122.0.
See all versions of @cotal-ai/cli for other recorded verdicts.
- Review the evidence and your use of @cotal-ai/cli@0.122.0 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.