Type Qdrant retrieval payloads

This commit is contained in:
Meta Vens 2026-05-19 13:00:08 +01:00
commit 91cff58858
4 changed files with 24 additions and 17 deletions

View file

@ -42,6 +42,7 @@ remembers doctrine; the branch remembers the blade.
- The native state ledger is closed typed state now: branch and evidence records no longer preserve flattened `extra` maps. Evidence entries are explicit `ts` / `type` / `status` / `note` / optional branch records.
- `epiphany-openai-codex-spine` no longer uses Codex's Responses client, model-provider layer, `codex-api`, `codex-model-provider`, `codex-model-provider-info`, or `codex-protocol` for outbound model requests. It reaches `codex-login` only through `epiphany-openai-auth-spine` so credential loading/refresh stays Codex-compatible while request/status/event/runtime surfaces stay typed Epiphany documents.
- `notes/epiphany-memory-graph-unified-plan.md` is the current memory architecture direction drawn from `Fractal Domains And The Cache That Bites` plus the VoidBot memory scar tissue. The repo semantic architecture/dataflow graph and agent memory are one `EpiphanyMemoryGraph` substrate with repo, role, short-term, incubation, agency, candidate-intervention, and evidence profiles. The first native skeleton is landed: `epiphany-state-model` owns shared graph document types, `epiphany-core::memory_graph` owns stable ids, validation, freshness, context cuts, and CultCache persistence, and `epiphany-memory-graph` can status/validate/context/smoke the store without Qdrant. Qdrant is a contribution cache for typed graph documents, not truth. `notes/repo-fractal-dataflow-cache-plan.md` and `notes/agent-memory-fractal-cache-plan.md` are profile plans, not rival organs.
- Retrieval's Qdrant boundary has a typed point payload now. Qdrant's HTTP requests, responses, and acknowledgements are JSON wire edges; Epiphany's retrieval facts (`path`, `line_start`, `line_end`, `excerpt`) are `QdrantPointPayload` before serialization/deserialization, not a loose payload map.
- Durable Epiphany state types now live in root crate `epiphany-state-model`. `codex-protocol` re-exports them only so old rollout/API callers keep compiling; do not add new durable `Epiphany*` state structs under `vendor/codex`. The deterministic `<epiphany_state>` prompt renderer also lives beside that typed model now, with prompt text under `epiphany-state-model/src/prompts/`; `epiphany-core` only re-exports it for compatibility. `epiphany-core` imports the native model directly and no longer depends on `codex-protocol`. Pressure derives from native `EpiphanyTokenUsageSnapshot`; Codex token telemetry is mapped at the bridge. Codex rollout/event reconstruction now lives in `codex-core::epiphany_rollout`, where the Codex `RolloutItem` / `EventMsg` types belong. Exact retrieval path matching is native to `epiphany-core`, so `epiphany-core` has no direct vendored Codex crate dependency.
- State-update contract proof and job launch/interrupt DTO ownership now live in native `epiphany-core`. The old giant `codex_thread.rs` Epiphany state-update test fixture was cut, and `epiphany-codex-bridge` no longer imports Epiphany job launch request types from `codex_core`; vendored `CodexThread` may call the native mutation/launch law for compatibility, but it must not regain ownership of state-update semantics, heartbeat launch-plan policy, or Epiphany DTOs.
- MCP remains JSON because MCP is a JSON protocol. The latest processor cut only moved MCP refresh/OAuth/status/resource/tool-call route handling into `codex_message_processor/mcp_routes.rs`, reducing the root processor to about 8,977 lines; it did not claim MCP payloads are CultNet or make MCP an Epiphany-owned runtime contract.