tnega@0.4.5
A minimal eval-first agent harness with spacetime-composable plugin lifecycles.
AI Security Review
scanned 1mo ago · by lpm-firewall-aiInspected behavior is consistent with an agent harness, including workspace approval rules and user-invoked execution tools. The self-dependency merits attention but this release lacks the automatic install hook required for the defined blocking chain, and no separate concrete attack was found.
Decision evidence
public snapshot- package.json has no preinstall, install, or postinstall hook; its runtime self-dependency therefore does not establish the specified malicious lifecycle chain.
- The flagged code in dist/bin.js, dist/cli-runtime.js, and dist/index.js reads workspace AGENTS.md for approval constraints without modifying it; approval defaults to manual.
- dist/tools.js provides request-driven execution and applies configured tool authorization policies.
- dist/web/assets/jszip.min-B3xE1mmA.js implements ZIP loading through its archive API.
- Web search selects a DeepSeek credential and sends it to the fixed DeepSeek API endpoint.
- The CLI activates agent execution through an explicit run command requiring a prompt.
Source & flagged code
16 flaggedFindings
Affected versions and remediation
This report applies to tnega@0.4.5.
See all versions of tnega for other recorded verdicts.
- Review the evidence and your use of tnega@0.4.5 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.