feat: add opt-in MCP decomposition hook points (design §10)

Adds the seven §10.2 hook-point modules + five env vars so Astrolabe Cloud can
offload document processing to the external document-processor / embedding
gateway. Purely additive: with every setting unset the server behaves exactly
as today, so self-hosters are unaffected (Deck #92).

Hook points (all default to current monolith behavior):
- config: EMBEDDING_PROVIDER, INGEST_MODE, STATUS_BACKEND,
  COLLECTION_METADATA_SOURCE, FACT_EVENT_EMITTER (+ supporting settings),
  validated in Settings.__post_init__ (fail-fast STATUS_BACKEND=local with
  INGEST_MODE=external); shared canonical.py.
- vector/payload_keys.py + acl_hash.py: cross-impl NAMESPACE/point_id (§2.2)
  and BLAKE2b-128 ACL hash (§11), pinned by fixtures shared with the
  document-processor repo.
- embedding/gateway_client.py: OpenAI-compatible GatewayProvider authenticating
  via M2M OIDC client-credentials (separate realm); manual-only registry entry.
- vector/collection_metadata.py: sentinel-point / API metadata source with env
  fallback.
- vector/queue/: hexagonal ingest producer ports + memory/NATS adapters
  (Postgres seam); INGEST_MODE=external publishes mcp.ingest.requested.{tenant}
  instead of the in-memory stream and skips the in-process processor pool. The
  lifespan becomes a composition root across both deployment branches.
- vector/queue/status.py: STATUS_BACKEND=bus subscriber feeding a StatusStore
  the vector-sync status endpoint reads.
- admin/payload_backfill.py: POST /api/v1/admin/payload-backfill (admin scope);
  processor writes the new payload keys; query-side ACL pre-filter gated behind
  ACL_PREFILTER_ENABLED (default off).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Chris Coutinho
2026-05-29 13:13:25 +02:00
co-authored by Claude Opus 4.8
parent c7da612f20
commit d883052fb8
37 changed files with 2534 additions and 55 deletions
@@ -50,6 +50,44 @@ class ProviderRegistry:
"""
settings = get_settings()
# 0. Gateway (manual-only; never autodetected). When
# EMBEDDING_PROVIDER=gateway, bypass the autodetect chain entirely
# and route embeddings through the Astrolabe Cloud embedding gateway
# (design §10.1/§10.2). Lazy import to avoid an import cycle
# (embedding/gateway_client → providers/openai → ... → registry).
if settings.embedding_provider == "gateway":
from ..embedding.gateway_client import ( # noqa: PLC0415
GatewayProvider,
GatewayTokenProvider,
)
# Settings.__post_init__ guarantees the URL is set when the
# provider is gateway; assert narrows the type for the checker.
assert settings.embedding_gateway_url is not None
# __post_init__ enforces all-or-nothing on the M2M creds, so
# checking one is enough to know the full triple is present.
token_provider = None
if settings.embedding_gateway_client_id:
assert settings.embedding_gateway_token_url is not None
assert settings.embedding_gateway_client_secret is not None
token_provider = GatewayTokenProvider(
token_url=settings.embedding_gateway_token_url,
client_id=settings.embedding_gateway_client_id,
client_secret=settings.embedding_gateway_client_secret,
scope=settings.embedding_gateway_scope,
)
logger.info(
"Using embedding gateway provider: url=%s, model=%s, auth=%s",
settings.embedding_gateway_url,
settings.embedding_gateway_model,
"oidc-m2m" if token_provider else "none",
)
return GatewayProvider(
base_url=settings.embedding_gateway_url,
embedding_model=settings.embedding_gateway_model,
token_provider=token_provider,
)
# 1. Bedrock
if (
settings.aws_region