Skip to content

bootstrap: implement first-pass relationship inference with provenance and confidence #307

Description

@flyingrobots

Summary

Implement the first-pass relationship inference layer for semantic bootstrap.

This covers Slices D and E from the Hill 1 semantic bootstrap spec.

Scope

Infer a small initial set of relationship types:

  • documents
  • references
  • touches
  • groups
  • conservative implements

And attach:

  • provenance metadata
  • rule-based confidence bands
  • enough surfaced information to review weak-confidence output

Acceptance Criteria

  • each inferred assertion includes inspectable provenance
  • confidence is rule-based and legible
  • low-confidence edges are clearly marked
  • the first slice improves repo understanding without requiring manual graph authoring first

Related

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    hill:h1-bootstrapHill 1 — zero-input semantic bootstraplane:foundationSupporting lane — core hardening, contracts, integrity

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions