Issues

• Change the wording around no issues found messaging

  runbot-hq/local-ai-code-review-action#103 · Filter by author eonist opened on Aug 22, 2026

• Empty reviews

  runbot-hq/local-ai-code-review-action#61 · Filter by author eonist opened on Aug 5, 2026

• Handle large PR inference timeouts with input chunking and adaptive output budgets

  runbot-hq/local-ai-code-review-action#58 · Filter by author eonist opened on Aug 4, 2026

• Time out error

  runbot-hq/local-ai-code-review-action#57 · Filter by author eonist opened on Aug 4, 2026
  1

• fix: think=true starves output tokens — add thinking-budget support and raise num_predict

  runbot-hq/local-ai-code-review-action#50 · Filter by author eonist opened on Aug 3, 2026

• Qwen still produces changelog/summary instead of review on large diffs (post-#45)

  runbot-hq/local-ai-code-review-action#49 · Filter by author eonist opened on Aug 3, 2026

• fix: add delimiter between few-shot example and diff to prevent model treating diff as example continuation

  runbot-hq/local-ai-code-review-action#48 · Filter by author eonist opened on Aug 3, 2026

• fix: instructions evicted from context window on large diffs — move to end of prompt

  runbot-hq/local-ai-code-review-action#47 · Filter by author eonist opened on Aug 3, 2026

• Prompt structuring tips

  runbot-hq/local-ai-code-review-action#46 · Filter by author eonist opened on Aug 2, 2026
  1

• feat: gate operational logs behind core.debug() for clean single-step review output

  enhancement
  runbot-hq/local-ai-code-review-action#43 · Filter by author eonist opened on Aug 2, 2026

• Add support for using AFM as the reviewer

  runbot-hq/local-ai-code-review-action#28 · Filter by author eonist opened on Aug 1, 2026

• Feature: Graph-aware context stuffing — inject targeted caller/callee context into LLM prompt via graphify

  enhancement
  runbot-hq/local-ai-code-review-action#20 · Filter by author eonist opened on Jul 25, 2026

• Feature: God Node Warning + Cross-File Link Awareness sections in PR reviews (fully static, no AI)

  enhancement
  runbot-hq/local-ai-code-review-action#19 · Filter by author eonist opened on Jul 25, 2026

• Feature: Add blast radius section to PR reviews using graphify community detection

  enhancement
  runbot-hq/local-ai-code-review-action#18 · Filter by author eonist opened on Jul 25, 2026

• Vision: how this action can match and beat CodeRabbit / Greptile

  runbot-hq/local-ai-code-review-action#17 · Filter by author eonist opened on Jul 25, 2026

• Design: 4th dimension — iterative (per-commit) vs preloaded (full PR) review timing

  runbot-hq/local-ai-code-review-action#16 · Filter by author eonist opened on Jul 25, 2026
  1

• Feature: RAG context injection via local Qdrant + Ollama embeddings

  runbot-hq/local-ai-code-review-action#15 · Filter by author eonist opened on Jul 25, 2026

• Research: competitor landscape and learnings for local 9B code review

  runbot-hq/local-ai-code-review-action#14 · Filter by author eonist opened on Jul 24, 2026

• Design: three-axis tension — CPU speed × model size × PR size

  runbot-hq/local-ai-code-review-action#13 · Filter by author eonist opened on Jul 24, 2026

• Feature: slice-and-aggregate review strategy — small (9B) vs normal (70B) mode

  runbot-hq/local-ai-code-review-action#12 · Filter by author eonist opened on Jul 24, 2026

• Fix: NON_CODE_PATTERNS filter not applied to diff block sent to model

  runbot-hq/local-ai-code-review-action#11 · Filter by author eonist opened on Jul 24, 2026

• Do research into how to make code review flows

  runbot-hq/local-ai-code-review-action#9 · Filter by author eonist opened on Jul 21, 2026

• Add some of the ideas around doing things in predefined steps etc

  runbot-hq/local-ai-code-review-action#8 · Filter by author eonist opened on Jul 21, 2026

• Add more meta data info in the ci steps

  runbot-hq/local-ai-code-review-action#7 · Filter by author eonist opened on Jul 18, 2026