AI risk level

Low

Recommendation

install

Model

~deepseek/deepseek-v4-flash-latest

Analyzed commit

75b56a3

Reviewed

1 month ago

The plugin is a well-structured AI command palette that communicates with a local daemon via harkctl. The code shows good security practices, such as passing API keys via stdin, sanitizing URLs before opening externally, and avoiding exposure of daemon stderr. The deterministic scan found no issues, and my review confirms no obvious malicious or destructive behavior.

• • The plugin spawns a bundled daemon (harkd) and communicates with it via a Unix socket; the daemon binary is not reviewed here, so its behavior is a potential risk, though it appears to be a legitimate AI assistant backend.
• • The plugin stores API keys in the system keyring via harkctl; while this is a standard practice, the security of the keyring integration depends on the daemon's implementation.
• • The plugin can capture screenshots and send them to AI providers; this is a privacy consideration, but it is a core feature and requires user action to trigger.

How this check works

This review combines the deterministic scan (the rule-based results above) with an independent look at the plugin's code by a language model. The model reads a trimmed sample of the repository's files, the manifest, and the README, then gives a plain-language risk level and a recommendation: install (no notable danger), review (look closer first), or avoid (clearly dangerous).

It runs on the same analyzed commit as the deterministic scan and is strictly advisory — it is not a security guarantee and never blocks a plugin by itself. A human moderator still reviews plugins before they are listed.

AI advisory only — automated analysis, not a security guarantee.