Skip to content

docs(specifications): add helix-db-feature-ideas spec#1603

Open
noahgift wants to merge 1 commit intomainfrom
docs/helix-db-feature-ideas
Open

docs(specifications): add helix-db-feature-ideas spec#1603
noahgift wants to merge 1 commit intomainfrom
docs/helix-db-feature-ideas

Conversation

@noahgift
Copy link
Copy Markdown
Contributor

Summary

Adds docs/specifications/helix-db-feature-ideas.md — a cross-pollination spec evaluating which helix-db patterns are worth adopting in aprender. Nine candidate proposals (HELIX-IDEA-001..009), each scoped with effort, target crate, non-goals, open questions, and acceptance signals.

Recommended (high priority):

  • HELIX-IDEA-001 — Persistent on-disk HNSW (extends in-memory aprender-core/src/index/hnsw.rs)
  • HELIX-IDEA-002 — Inventory-based MCP handler auto-registration (replaces hardcoded Vec at aprender-mcp/src/server.rs:221-233)
  • HELIX-IDEA-005 — Hybrid retrieval (BM25 + dense); the RAG spec lists this as a design goal but no implementation exists
  • HELIX-IDEA-006 — Reranking pipeline (RRF / MMR / cross-encoder)
  • HELIX-IDEA-007 — Snapshot / atomic backup primitive
  • HELIX-IDEA-009 — Constant-time API-key auth for apr serve

Speculative / deferred:

  • HELIX-IDEA-003 — Compile-time-validated DSL macro pattern (needs concrete pain point first)
  • HELIX-IDEA-004 — Multi-target deployment scaffolding (deferred until hosted-inference direction is set)
  • HELIX-IDEA-008 — Schema versioning macro (defer until first painful registry migration)

Explicitly rejected (with reasons in §3): LMDB storage swap, HelixQL the language, helix-db graph traversal model, embedding-provider abstraction, browser dashboard, vendor-specific metrics pipeline, heed3 for registry (already SQLite).

What this is and isn't

  • This is an ideas / proposal document, not an implementation. No code changes.
  • §1.3 grounds claims in verified facts about aprender's current state (HNSW LOC, registry storage, MCP wiring, redb availability).
  • §6 contains a falsification log: the initial draft incorrectly claimed MCP handler discovery was contracts-mediated; corrected to "contracts mediate schema, manual hardcoded Vec mediates discovery." That correction sharpens HELIX-IDEA-002's value.

Test plan

  • Markdown renders cleanly on GitHub
  • All file paths cited in the spec resolve in the current main (e.g., crates/aprender-core/src/index/hnsw.rs, crates/aprender-mcp/src/server.rs:221-233)
  • No code changes — ci / gate and workspace-test should pass without rebuild impact
  • Reviewers should sanity-check effort estimates and challenge the "rejected" reasoning in §3

🤖 Generated with Claude Code

Cross-pollination spec evaluating helix-db patterns for adoption in
aprender. Nine candidates (HELIX-IDEA-001..009) covering persistent
HNSW, inventory-based MCP handler registration, compile-time DSL
macro pattern, multi-target deployment, hybrid retrieval (BM25 +
dense), reranking pipeline (RRF/MMR/cross-encoder), snapshot/backup,
schema migration macro, and constant-time API-key auth for apr serve.

Each proposal scoped with effort, target crate, non-goals, open
questions, and acceptance signals. Section 1.3 grounds the spec in
verified facts about aprender's current state; section 6 logs one
falsified-and-corrected claim from the initial draft (MCP handler
discovery is hardcoded, not contracts-mediated).

Section 3 enumerates rejected candidates (LMDB swap, HelixQL the
language, embedding-provider abstraction, browser dashboard,
vendor-specific metrics) with explicit reasoning.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant