Skip to content

[backlog] Extend policy model to resources and prompts #26

Description

@ernestprovo23

Surfaced via a conclave council run (synthesize + adversarial passes), 2026-06-07.

Council verdict: CUT this cycle (revisitable). Directionally correct — resource URIs and prompt templates are real MCP-CAPSURF vectors — but the structural diff engine (#15) does not yet cover these artifact types, and the project explicitly excludes broad fuzzy NLP scanning. Extending policy coverage without the corresponding structural-diff semantics produces a half-measure.

Revisit when: #15 is proven AND the schema-diff approach generalizes cleanly to resource/prompt schemas.
Status: parked backlog — not active work. Labeled wontfix for this cycle, not permanently.

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or requestroadmapPlanned work from the threat model / roadmapwontfixThis will not be worked on

    Type

    No type

    Fields

    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