registry / @bacnh85/pi-serena / 0.7.0

@bacnh85/pi-serena@0.7.0

Pi extension that provides Serena semantic code tools through a persistent worker.

AI Security Review

scanned 21h ago · by lpm-firewall-ai

LPM treats this as warn-only first-party agent extension lifecycle risk. A Pi extension loads its entrypoint and registers semantic-code, editing, memory, and lifecycle hooks. On a user tool request it starts a local Serena Python bridge; strict tool blocking is explicitly environment-gated.

Static reason

No blocking static signals were detected.

Trigger

Pi loads the extension; Serena actions require user/agent tool calls, except optional `SERENA_EAGER_STARTUP=1`.

Impact

Can influence Pi tool workflow and perform Serena-provided project mutations when its registered tools are invoked.

Mechanism

Pi extension lifecycle hooks plus local Serena subprocess bridge

Rationale

Source inspection found no install hook or concrete malicious behavior. The Pi lifecycle hooks and local subprocess capability warrant a non-blocking warning under the extension-risk policy.

Evidence

package.jsonextensions/index.tsextensions/worker.tsextensions/lib/detect.tsextensions/lib/guidance.ts

Decision evidence

public snapshot

AI called this Suspicious at 86.0% confidence as Dangerous Capability with low false-positive risk.

Evidence for warning

`extensions/index.ts` registers Pi lifecycle hooks and can block tool calls when an opt-in env flag is set.

@bacnh85/pi-serena@0.7.0

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Source & flagged code