Issues

• answer(): send the whole document when it fits within max_context_tokens

  enhancement
  mrutunjay-kinagi/ragsearch#120 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.3.0

• Unstructured files: indexed text includes metadata/path/parser columns and NUL characters

  bug
  mrutunjay-kinagi/ragsearch#119 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.2.1

• Retrieve a wider candidate pool before reranking (candidate_k)

  enhancement
  mrutunjay-kinagi/ragsearch#116 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.3.0

• Open issue creation to the public, with a bug-report template

  enhancement
  mrutunjay-kinagi/ragsearch#115 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.2.1

• Refresh the Stage 0 epic (#84)

  documentation
  mrutunjay-kinagi/ragsearch#114 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.2.1

• Add ROADMAP.md

  documentation
  mrutunjay-kinagi/ragsearch#113 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.2.1

• Rewrite the dataset analytics cookbook and notebook

  documentation
  mrutunjay-kinagi/ragsearch#112 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.2.1

• Deploy the Sphinx docs site automatically

  documentation
  mrutunjay-kinagi/ragsearch#111 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.2.1

• Publish to PyPI with Trusted Publishing instead of an API token

  enhancement
  mrutunjay-kinagi/ragsearch#110 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.2.1

• Support the official LiteParse CLI (@llamaindex/liteparse) as an optional parser (0.3.0)

  enhancement
  mrutunjay-kinagi/ragsearch#102 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.3.0

• Web server: blocking behaviour, script-exit HTTP 500s, and production (WSGI) support for run()

  bug
  mrutunjay-kinagi/ragsearch#98 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.3.0

• Document the embedding-model identity string as a cache compatibility contract

  documentation
  mrutunjay-kinagi/ragsearch#97 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026

• Embedding cache identity for custom providers without a model name

  enhancement
  mrutunjay-kinagi/ragsearch#96 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026

• Rename the setup implementation module to _setup.py; deprecate the ragsearch.setup module path (before 1.0)

  enhancement
  mrutunjay-kinagi/ragsearch#94 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026

• Use a query-specific embedding (Cohere input_type="search_query") for search queries

  enhancement
  mrutunjay-kinagi/ragsearch#92 · Filter by author mrutunjay-kinagi opened on Sep 29, 2026 · 0.3.0
  2

• [EPIC] Stage 0: P0 correctness fixes

  enhancement
  mrutunjay-kinagi/ragsearch#84 · Filter by author mrutunjay-kinagi opened on Sep 23, 2026

• [P0-9] Parsing gaps: DOCX tables dropped, scanned PDFs yield no text

  bug
  mrutunjay-kinagi/ragsearch#83 · Filter by author mrutunjay-kinagi opened on Sep 23, 2026 · 0.3.0

• [P0-5] Hygiene: root logging config, required API key for local providers, 0.0.0.0 bind

  bug
  mrutunjay-kinagi/ragsearch#79 · Filter by author mrutunjay-kinagi opened on Sep 23, 2026 · 0.2.1
  1

• [P0-4] Numeric columns excluded from structured combined_text

  bug
  mrutunjay-kinagi/ragsearch#78 · Filter by author mrutunjay-kinagi opened on Sep 23, 2026 · 0.3.0

• [P0-2] ChromaDB mode: search() and answer() always fail

  bug
  mrutunjay-kinagi/ragsearch#76 · Filter by author mrutunjay-kinagi opened on Sep 23, 2026 · 0.3.0
  1

• Product use case

  mrutunjay-kinagi/ragsearch#9 · Filter by author tusharchou opened on Aug 11, 2025