Skip to content

docs(claude): fix Embedding-Configuration SSoT — JSON-File, nicht Neo4j (#958) - #959

Merged
arn0ld87 merged 2 commits into
mainfrom
docs/embedding-config-neo4j-ssot-fix
Jul 28, 2026
Merged

docs(claude): fix Embedding-Configuration SSoT — JSON-File, nicht Neo4j (#958)#959
arn0ld87 merged 2 commits into
mainfrom
docs/embedding-config-neo4j-ssot-fix

Conversation

@arn0ld87

Copy link
Copy Markdown
Owner

Kontext

Behebt die SSoT-Discrepancy aus #958, die bei der Codex-Findings-Verifikation an PR #957 zutage trat.

Befund

Der Agora-spezifische CLAUDE.md-Abschnitt „Embedding-Configuration (ADR-0007)" behauptete:

Aktive Konfiguration lebt in Neo4j (EmbeddingConfiguration-Knoten, eindeutig pro Provider/Model/Dim).

Das ist falsch. Code ist führend:

ADR-0007 selbst ist korrekt — es beschreibt genau diesen Store (flock, atomares Write, 0600). Auch docs/agents/architecture-ssot.md nennt nur die Services und ist nicht betroffen. Nur der CLAUDE.md-Absatz war gedriftet.

Änderung

Einzeilige Korrektur des Absatzes auf EmbeddingConfigurationStore (JSON-Datei), mit Verweisen auf embedding_contract.py, embedding_configurations/service.py und …/legacy.py. Migrations-Lifecycle- und Gemini-Aussagen unverändert (waren korrekt).

Warum Option A (Code kanonisch)

Option B (Code nach Neo4j migrieren) ist nicht tragfähig: ADR-0007 selbst spezifiziert den JSON-Store als Umsetzung für Slice 4.2. Die JSON-Persistenz ist bewusst so designed (atomic, secret-frei, file-locked). Neo4j-Persistenz war nie intendiert — die CLAUDE.md-Aussage war reine Doku-Drift.

Verifikation

  • Pre-Push-Gate schemas: green (46 Schemas, Version 0.8.0, STATUS synchron).
  • 1 file changed, 1 insertion(+), 1 deletion(-).

Closes #958.

CLAUDE.md-Abschnitt "Embedding-Configuration (ADR-0007)" behauptete
fälschlich, aktive Konfiguration lebe in Neo4j (EmbeddingConfiguration-
Knoten). Code (embedding_configuration_store.py) persistiert aber in
einer flachen JSON-Datei unter AGORA_DATA_DIR/ mit flock + atomarem
Write (os.replace) + Datei-Modus 0600. ADR-0007 selbst beschreibt genau
diesen Store korrekt — nur der CLAUDE.md-Absatz war gedriftet.

Korrigiert auf EmbeddingConfigurationStore (JSON-Datei), Verweis auf
embedding_contract.py (Pydantic-Vertrag, kein Neo4j-Knoten) sowie
embedding_configurations/service.py + legacy.py. Migrations-Lifecycle-
und Gemini-Aussagen unverändert (waren korrekt).

Refs #958

Co-Authored-By: Claude <noreply@anthropic.com>
@coderabbitai

coderabbitai Bot commented Jul 28, 2026

Copy link
Copy Markdown
Contributor

Warning

Review limit reached

@arn0ld87, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 6 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details
⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 170b7b79-4b9b-4779-bf5b-80fbbc954cb4

📥 Commits

Reviewing files that changed from the base of the PR and between 7fcc66a and 6d8955d.

📒 Files selected for processing (1)
  • CLAUDE.md

Comment @coderabbitai help to get the list of available commands.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 2837de6bab

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread CLAUDE.md
## Embedding-Configuration (ADR-0007)

Aktive Konfiguration lebt in Neo4j (`EmbeddingConfiguration`-Knoten, eindeutig pro Provider/Model/Dim). Lese-/Schreibpfade ausschließlich über `backend/app/services/embedding_service.py` und `embedding_migration.py`. Bei Modell-Wechsel: Migrations-Lifecycle `pending → running → validating → completed | failed | rolled_back`. Gemini-Re-Embedding ist explizit „noch nicht unterstützt“ — nicht vortäuschen.
Aktive Konfiguration wird vom [`EmbeddingConfigurationStore`](backend/app/services/embedding_configuration_store.py) in einer flachen JSON-Datei unter `AGORA_DATA_DIR/embedding_configurations.json` persistiert (`flock`-Sperre, atomares Write via `os.replace`, Datei-Modus 0600) — **nicht** in Neo4j. Index-Versionen liegen in einer Sibling-JSON (`embedding_index_versions.json`). `EmbeddingConfiguration` ist ein Pydantic-Vertrag ([`backend/app/contracts/embedding_contract.py`](backend/app/contracts/embedding_contract.py)), kein Neo4j-Knoten. Lese-/Schreibpfade über `backend/app/services/embedding_service.py` und `embedding_migration.py`; Service-Logik in `backend/app/services/embedding_configurations/service.py`, Legacy-Read-Only-Adapter in `…/legacy.py`. Bei Modell-Wechsel: Migrations-Lifecycle `pending → running → validating → completed | failed | rolled_back`. Gemini-Re-Embedding ist explizit „noch nicht unterstützt“ — nicht vortäuschen.

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Point the embedding link at an existing canonical module

When an agent follows the newly linked read/write path, it resolves to backend/app/services/embedding_service.py, which does not exist; the embedding implementation is at backend/app/storage/embedding_service.py, while configuration persistence is handled by embedding_configuration_store.py and embedding_configurations/service.py. This makes the SSoT guidance lead to a dead path, so the link and wording should identify the appropriate existing module.

Useful? React with 👍 / 👎.

@arn0ld87
arn0ld87 merged commit 29d0233 into main Jul 28, 2026
28 checks passed
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.

docs: SSoT-Discrepancy Embedding-Configuration — Code persistiert in JSON-Datei, Docs behaupten Neo4j

1 participant