Skip to content

Agent task: mature Holmes language intelligence fabric to M2/M3 product spine #4

@mdheller

Description

@mdheller

Agent-directed work package.

Target agent: GitHub Copilot coding agent; Codex may also take this if enabled via GitHub Agent HQ.

Context:
Holmes is the SocioProphet language intelligence fabric: governed classical NLP, neural NLP, semantic search, retrieval, knowledge graphs, foundation-language services, guardrails, evals, and investigative agentic discovery. Sherlock Search is the discovery engine; Holmes is the detective layer above it.

Scope:

  1. Inspect the live repository before editing.
  2. Add/repair product README and docs/ARCHITECTURE.md if incomplete.
  3. Add repo.maturity.yaml conforming to repo-maturity.v1 semantics from SocioProphet/functional-model-surfaces.
  4. Add contracts/holmes-surface.v1.json or equivalent compact product-surface manifest.
  5. Add examples/holmes.surface.example.json.
  6. Add Makefile validation target and minimal validator tooling.
  7. Define Holmes boundaries: it may orchestrate Sherlock Search, NLP labs, embedding/rerank, graph conversion, guardrails, and agent tools, but it must not own model promotion, mutable model lifecycle, or SourceOS OS state.

Acceptance criteria:

  • make validate passes.
  • Product docs clearly explain Holmes, Sherlock Search, 221B, Mycroft, Moriarty Bench, Irene Shield, The Canon, and Deduction Engine as product roles.
  • repo.maturity.yaml marks current maturity honestly and lists next actions.
  • Future Prophet CLI work can consume the product-surface manifest without inventing semantics.

Boundary:
One PR only. Do not touch unrelated repositories. Do not add model weights, datasets, secrets, or runtime credentials.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions