@telora/daemon@0.22.189
Agent orchestration daemon for Telora - spawns and manages Claude Code instances
AI Security Review
scanned 3d ago · by lpm-firewall-aiSource inspection found first-party agent configuration setup that affects later agent sessions, which warrants a warning under the agent integration policy. The manifest's prepublishOnly command builds the package, and the inspected runtime operations do not establish a publish-blocking attack.
Decision evidence
public snapshot- The manifest defines prepublishOnly as a build and metadata-stamping command, with no install-time hook.
- The connect command provisions agent scaffolding and MCP registrations in project or agent configuration.
- The Claude connector merges Telora settings into existing configuration and writes the result when it changes.
- The connect flow can also register workstation tracking and merge hooks into user Claude settings when telora-ai is available.
- The inspected worktree and shell operations concern dependency installation analysis and daemon configuration reloads; they do not establish a package-install attack.
- The configuration setup is part of the explicit connect flow, and the manifest's lifecycle command is prepublishOnly.
- The cited runtime operations show worktree dependency installation, install-script analysis, and configuration reload behavior rather than credential transfer or destructive actions.
Source & flagged code
6 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@0.22.189.
See all versions of @telora/daemon for other recorded verdicts.
- Review the evidence and your use of @telora/daemon@0.22.189 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.