@cotal-ai/cli@0.133.13
Cotal mesh CLI: up, join, watch, console, spawn, mint, channels, history, ext.
AI Security Review
scanned 3d ago · by lpm-firewall-aiThe source shows disclosed, default-on runtime telemetry with an opt-out, but the notice is only shown in TTY sessions, making telemetry undisclosed for noninteractive runs. This warrants a warning; the flagged agent-skill writes are tied to explicit setup actions and show no separate attack behavior.
Decision evidence
public snapshot- When telemetry settings are first created, the dependency defaults to the regular level and generates an install ID.
- The CLI arms telemetry for normal commands; it skips sending only when telemetry is off or the command is marked silent.
- The telemetry notice is printed only when stderr is a TTY, so noninteractive command runs can send without displaying it.
- Regular command events include a hashed account identifier, command and flag names, and scrubbed error details.
- Telemetry is posted to https://cotal.ai/v1/telemetry when the process exits.
- The CLI offers `cotal telemetry off` to disable telemetry.
- The telemetry is described as anonymous; its source says it excludes argument values, paths, names and content.
- The endpoint is the package publisher's own service, and the source describes scrubbing paths and identifying values from error text.
- The separate agent-skills reconciliation runs through the explicit `cotal setup` command and manages the package's authored skill files.
- No npm lifecycle scripts are declared, and no install-time behavior was identified.
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@0.133.13.
See all versions of @cotal-ai/cli for other recorded verdicts.
- Review the evidence and your use of @cotal-ai/cli@0.133.13 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.