boost-backend — RAG knowledge pipeline backend (issue 9 of 15)

[gabemontero]

Labels: ready-to-code
Depends on: Issue 3

Implement the RAG knowledge pipeline backend: document ingestion, vector store integration, content hash tracking, and RAG query endpoints.

Tasks

From openspec/changes/platform-operations-deployment/tasks.md:

• 1.4 DocumentSyncService content hashes: cacheService with long TTL

Specifications

• openspec/changes/ai-chat-interaction-experience/specs/rag-knowledge/spec.md
• openspec/changes/platform-operations-deployment/specs/rag-pipelines/spec.md
• openspec/changes/platform-operations-deployment/specs/cache-migration/spec.md — DocumentSyncService scenario

[fullsend-ai-triage]

Triage: Blocked

This issue is blocked by #3299 (Runtime configuration engine with Zod validation), which is still open.

The RAG knowledge pipeline — specifically the DocumentSyncService content hash tracking via cacheService — depends on the runtime configuration infrastructure being established first. Until #3299 lands, work on this issue cannot meaningfully proceed.

Additionally, the specification files referenced in the issue body (under openspec/) do not appear to exist in the repository yet, which may indicate further prerequisite work is needed.

This issue will be re-evaluated once the blocking dependency is resolved.

---

Labels: This is a boost-backend feature issue in the AI integrations domain, matching the ai-integrations label per repo conventions.

[gabemontero]

/fs-code

[gabemontero]

/fs-code --force

[fullsend-ai-coder]

🤖 Finished Code · ✅ Success · Started 4:20 PM UTC · Completed 4:36 PM UTC
Commit: a50b883 · View workflow run →