@telora/daemon@0.23.57
Agent orchestration daemon for Telora - spawns and manages Claude Code instances
AI Security Review
scanned 9d ago · by lpm-firewall-aiSource inspection found no malicious behavior. The flagged lifecycle, shell, and package-manager signals correspond to a publish-time build script, Git verification and update operations, and worktree provisioning; agent hook configuration is reached through connect setup.
Decision evidence
public snapshot- The manifest exposes the daemon CLI, has only a prepublishOnly build-and-stamp lifecycle script, and has no dependency on itself.
- The connect flow performs workstation tracking only when telora-ai is found; it merges Telora hook entries into the user's Claude settings while preserving existing configuration.
- The configuration writers create or merge Telora-related files and ignore entries for managed products; the worktree hook writes a scoped read-only deny overlay.
- Worktree npm ci runs in a selected package directory, while Lacuna setup requires a configured source before building an install plan.
- Self-update code uses its detected install mode, and Git-source updates pull and install within the detected checkout.
- The flagged verification shell operation reads a Git object, and the audit phase limits its assessor to an allowlisted set of Telora MCP tools.
Source & flagged code
8 flagged · loading sourcePackage source references child process execution.
dist/update/self-update.jsView on unpkg · L27Package source invokes a package manager install command at runtime.
dist/update/self-update.jsView on unpkg · L17Findings
Affected versions and remediation
This report applies to @telora/daemon@0.23.57.
See all versions of @telora/daemon for other recorded verdicts.
- Review the evidence and your use of @telora/daemon@0.23.57 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.