Skip to content

AST-based static extraction for skill packs from untrusted registries #279

@tps-anvil

Description

@tps-anvil

Context

Sherlock flagged during review of feat-skill-add-pack (PR #278): dynamic import() of dist/index.js runs top-level code before the Flair scanner sees content. This is an inherent trust issue — if packs come from untrusted registries, malicious code in dist/index.js executes during import.

Current state

The scanner runs after import, so it can't inspect the module source before execution. For npm's trust model this is accepted risk (npm packages already run arbitrary code on install).

Proposed

AST-based static extraction: parse dist/index.js as source text, extract export const ruleNames/rules/skillSummary declarations without executing the module. This would let scanner run before any code executes.

Priority

Low — not blocking for trusted registries. Would become blocking if TPS ever supports community/untrusted registries.

Related

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions