Spaces:

amer003100
/

vectorless_RAG

Running

App Files Files Community

vectorless_RAG

Commit History

Oct 27 Today

10:42
feat(documents): add bulk DELETE /documents endpoint and active task cancellation Running

amer003100

9f0929b
10:31
feat(rag): isolate conversational mode from stored docs unless explicitly attached or mentioned

amer003100

2926b6d
10:27
fix(chat): detect casual conversation to prevent unwanted web search on chit-chat

amer003100

a7887b4
10:16
test(auth): update client isolation tests to verify graceful fallback

amer003100

2b0613b
10:13
fix(auth): fallback to default client ID gracefully and expose shared documents

amer003100

bbcb666
09:47
fix(postgres): safely defer owner_id index creation to auto-migration step

amer003100

0deb3a0
09:35
fix(rag): prioritize document tree root title and exclude generic headings in research queries

amer003100

1fa8476
09:34
feat(research): enrich academic queries with active document topic and enforce verified arXiv citations

amer003100

29a9e61
09:23
fix(llm): guard Groq fallback against 8000 TPM limit with prompt compaction and context trimming

amer003100

2789a91
09:21
fix(cors): permit Vercel and HF Spaces domains via origin regex and update tests

amer003100

9257463
09:08
fix(studio): switch Groq model to high-quota gpt-oss-120b and parallelize quiz generation

amer003100

4677a3b
08:48
feat(studio): enrich mindmap nodes with full text and provide evidence_text excerpts in quiz

amer003100

f3f0bbe
07:53
fix(postgres): normalize DSN credentials and URL-encode special characters in password

amer003100

698fc89
07:43
chore(deps): declare missing runtime dependencies and refresh requirements lock

amer003100

841b61d
07:26
feat(llm): add explicit client and streaming timeouts with failover preservation

amer003100

ee92077
07:21
feat(cors): configure restricted CORS allowlist from environment

amer003100

236f036
07:20
feat(upload): harden PDF upload with chunked size limits, magic bytes check, and sanitized storage

amer003100

45a0af7
07:18
feat(security): implement SSRF protection, safe manual redirects, and body limits in web reader

amer003100

699abca
07:15
feat(ratelimit): implement dual-key in-memory sliding window rate limiter

amer003100

efdcc61
07:10
feat(auth): enforce client ownership and data isolation across all endpoints and databases

amer003100

a506cc3
06:27
fix(startup): ensure static and data directories exist before mounting and writing

AI Assistant

92742e1
05:17
docs: accuracy pass on README (fix latency claims, subsystem count, units, links, and naming)

AI Assistant

f7154a3
05:06
docs: update main README with nucleus top_p architecture, fastmcp suite, study studio, and 95-question benchmark tables

AI Assistant

723adc4
04:57
feat(rag): implement nucleus top_p retrieval and exact inner page citation pinning⁠

AI Assistant

de0d0d6
03:40
feat(eval): add unified FinanceBench benchmark suite and full-system verification report

AI Assistant

8b4bcb4
03:20
feat: upgrade FastMCP and internal tools to Research & Study MCP suite⁠

AI Assistant

dad455c
02:53
feat(studio): implement Phase 4 Interactive Study Studio backend with mindmap graph and AI quiz generator

AI Assistant

9ad6c13
02:36
feat(retrieval): implement Phase 3 query intent classification and dossier root context injection

AI Assistant

f4d3bb7
02:23
feat(retrieval): implement vectorless hybrid search with zero-dependency BM25 and weighted RRF

AI Assistant

3a2ce42
02:09
feat(rag): inject page boundaries and implement sub-node paragraph pruning for precise citations

AI Assistant

2bc0095

Oct 18 9 days ago

00:50
docs: add quantitative evaluation benchmarks (80-question single-doc test) and report

AI Assistant

b3242a4

Oct 17 10 days ago

23:11
feat: add active HTTP link testing and strict content validation prompt directives

AI Assistant

7d5d266
23:07
docs: rewrite README with full architecture, FastMCP, and PostgreSQL details

AI Assistant

fcead5f
22:48
feat: add FastMCP web search server, tool and dual citations

AI Assistant

63f56d4
21:16
feat: expose database connection diagnostics on /health

AI Assistant

190fb26
21:09
chore: trigger HF Space restart to load DATABASE_URL secret

AI Assistant

4e3a9ba
20:54
feat: integrate PostgreSQL persistence for documents, trees, and chat sessions

AI Assistant

8eaad14

Oct 15 12 days ago

04:10
fix: ensure chat existence in persistent DB, safe history endpoint, and data mount support

AI Assistant

a57411f
02:28
docs: update README with multi-document Vectorless RAG, memory architecture, and session isolation

AI Assistant

c35d9d3
02:26
feat: add multi-document vectorless RAG, session isolation, and memory services

AI Assistant

0039ed6
01:38
docs: update README with Atlas Document Intelligence API and Groq fallback details

AI Assistant

56cf8f7
01:21
feat: add Atlas identity, document endpoints, and Groq fallback for depleted tokens

AI Assistant

dff6fcd

Oct 5 22 days ago

19:37
feat: asynchronous PDF upload and background indexing with status polling

AI Assistant

d13ebb9
18:47
chore: update Dockerfile to python 3.11, set HF app_port, and ignore log files

AI Assistant

1b81d8f
18:24
update

AI Assistant

be59b3a
17:41
feat: add RAG application modules, services, routers, and schemas for document and chat processing

AI Assistant

7b5352e

May 13

15:36
Enhance LLMService to coerce node IDs and improve tree normalization

AI Assistant

9c36dc8
15:32
Update README.md to enhance description and clarify multimodal RAG support

AI Assistant

b516520
15:29
update

AI Assistant

9b5a2e4
15:09
Refactor LLMService to integrate gemini client and update dependencies in pyproject.toml

AI Assistant

ad67d8e

Commit History

Oct 27 Today

feat(documents): add bulk DELETE /documents endpoint and active task cancellation Running

feat(rag): isolate conversational mode from stored docs unless explicitly attached or mentioned

fix(chat): detect casual conversation to prevent unwanted web search on chit-chat

test(auth): update client isolation tests to verify graceful fallback

fix(auth): fallback to default client ID gracefully and expose shared documents

fix(postgres): safely defer owner_id index creation to auto-migration step

fix(rag): prioritize document tree root title and exclude generic headings in research queries

feat(research): enrich academic queries with active document topic and enforce verified arXiv citations

fix(llm): guard Groq fallback against 8000 TPM limit with prompt compaction and context trimming

fix(cors): permit Vercel and HF Spaces domains via origin regex and update tests

fix(studio): switch Groq model to high-quota gpt-oss-120b and parallelize quiz generation

feat(studio): enrich mindmap nodes with full text and provide evidence_text excerpts in quiz

fix(postgres): normalize DSN credentials and URL-encode special characters in password

chore(deps): declare missing runtime dependencies and refresh requirements lock

feat(llm): add explicit client and streaming timeouts with failover preservation

feat(cors): configure restricted CORS allowlist from environment

feat(upload): harden PDF upload with chunked size limits, magic bytes check, and sanitized storage

feat(security): implement SSRF protection, safe manual redirects, and body limits in web reader

feat(ratelimit): implement dual-key in-memory sliding window rate limiter

feat(auth): enforce client ownership and data isolation across all endpoints and databases

fix(startup): ensure static and data directories exist before mounting and writing

docs: accuracy pass on README (fix latency claims, subsystem count, units, links, and naming)

docs: update main README with nucleus top_p architecture, fastmcp suite, study studio, and 95-question benchmark tables

feat(rag): implement nucleus top_p retrieval and exact inner page citation pinning⁠

feat(eval): add unified FinanceBench benchmark suite and full-system verification report

feat: upgrade FastMCP and internal tools to Research & Study MCP suite⁠

feat(studio): implement Phase 4 Interactive Study Studio backend with mindmap graph and AI quiz generator

feat(retrieval): implement Phase 3 query intent classification and dossier root context injection

feat(retrieval): implement vectorless hybrid search with zero-dependency BM25 and weighted RRF

feat(rag): inject page boundaries and implement sub-node paragraph pruning for precise citations

Oct 18 9 days ago

docs: add quantitative evaluation benchmarks (80-question single-doc test) and report

Oct 17 10 days ago

feat: add active HTTP link testing and strict content validation prompt directives

docs: rewrite README with full architecture, FastMCP, and PostgreSQL details

feat: add FastMCP web search server, tool and dual citations

feat: expose database connection diagnostics on /health

chore: trigger HF Space restart to load DATABASE_URL secret

feat: integrate PostgreSQL persistence for documents, trees, and chat sessions

Oct 15 12 days ago

fix: ensure chat existence in persistent DB, safe history endpoint, and data mount support

docs: update README with multi-document Vectorless RAG, memory architecture, and session isolation

feat: add multi-document vectorless RAG, session isolation, and memory services

docs: update README with Atlas Document Intelligence API and Groq fallback details

feat: add Atlas identity, document endpoints, and Groq fallback for depleted tokens

Oct 5 22 days ago

feat: asynchronous PDF upload and background indexing with status polling

chore: update Dockerfile to python 3.11, set HF app_port, and ignore log files

update

feat: add RAG application modules, services, routers, and schemas for document and chat processing

May 13

Enhance LLMService to coerce node IDs and improve tree normalization

Update README.md to enhance description and clarify multimodal RAG support

update

Refactor LLMService to integrate gemini client and update dependencies in pyproject.toml