@telora/daemon@0.22.4
Agent orchestration daemon for Telora - spawns and manages Claude Code instances
AI Security Review
scanned 3d ago · by lpm-firewall-aiThe source establishes a default-off, operator-configured sidecar installation and execution capability, which warrants a warning for its residual extension lifecycle risk. The package manifest has no install lifecycle hook, and the inspected runtime receiver is ordinary promise cleanup.
Decision evidence
public snapshot- Daemon startup starts the Lacuna sidecar only when enabled; its source comes from operator configuration and can be a Git URL or local path.
- If the sidecar environment is missing, the supervisor runs an install plan that creates a Python virtual environment and installs from the configured checkout.
- The unresolved lifecycle receiver is a promise .finally() that resolves the merge result; it does not perform an account action.
- The package has no preinstall, install, or postinstall hook; its prepublishOnly script builds and stamps build information.
- The connect and IDE integration excerpts describe Telora configuration and MCP registration, while the audit hook excerpt describes a write-deny overlay.
- The sidecar is opt-in, and failed startup is handled without preventing the daemon from running.
- The reviewed configuration selects the Lacuna source from environment or config values rather than a publisher-selected source.
Source & flagged code
6 flagged · loading sourcePackage source references child process execution.
dist/update/self-update.jsView on unpkg · L25Package source invokes a package manager install command at runtime.
dist/update/self-update.jsView on unpkg · L16Findings
Affected versions and remediation
This report applies to @telora/daemon@0.22.4.
See all versions of @telora/daemon for other recorded verdicts.
- Review the evidence and your use of @telora/daemon@0.22.4 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.