deepset-ai/haystack

View on GitHub

Open-source AI orchestration framework for building context-engineered, production-ready LLM applications. Design modular pipelines and agent workflows with explicit control over retrieval, routing, memory, and generation. Built for scalable agents, RAG, multimodal applications, semantic search, and conversational systems.

Stars: 26.6k
Forks: 3.2k
Open beginner issues: 0
Indexed issues: 92
Avg merge: 1d 3h
Merged PRs (30d): 194

Dominant language: Python
License: Apache-2.0
Last GitHub push: Sep 18, 2026
Latest indexed: Sep 20, 2026
Contributing guide: Contributing guide
Code of conduct: Code of conduct
Beginner labels: No beginner labels indexed

Beginner friendly Everything

Open All

92 open issues indexed

FileSystemSkillStore advertises escaping symlinks as readable bundled files Open

deepset-ai/haystack#12820 · 1 assignee · Opened Sep 20, 2026
Fix async LinkContentFetcher custom client headers Open

deepset-ai/haystack#12817 · 1 assignee · Opened Sep 19, 2026
FilterPolicy.MERGE unions OR/NOT filters instead of intersecting them, returning documents the init filter excludes Open

deepset-ai/haystack#12815 · 1 assignee · Opened Sep 19, 2026
inputs_from_state typos pass construction when a tool takes no input Open

deepset-ai/haystack#12811 · 1 assignee · Opened Sep 19, 2026
deserialize_secrets_inplace with recursive=True stops restoring the keys you listed Open

deepset-ai/haystack#12809 · 1 assignee · Opened Sep 19, 2026
InMemoryDocumentStore.embedding_retrieval() ignores the store-level return_embedding default Open

Difficulty 2/5 1-3 hours Newbie friendliness 82/100

deepset-ai/haystack#12803 · 2 comments · Opened Sep 18, 2026
Integration proposal: a memory store with verified transfer between pipelines and deployments Open

Difficulty 5/5 Over a week Newbie friendliness 25/100

deepset-ai/haystack#12802 · Opened Sep 18, 2026
CSVToDocument row mode overwrites a CSV column named row_number Open

deepset-ai/haystack#12787 · 1 assignee · Opened Sep 17, 2026
CSVDocumentCleaner deletes a row that says N/A, and both CSV components empty those cells Open

Difficulty 2/5 1-3 hours Newbie friendliness 88/100

deepset-ai/haystack#12784 · 2 comments · Opened Sep 16, 2026
Agent generation_kwargs["tools"] replaces agent.tools instead of being treated as client-side specs Open

Difficulty 4/5 3-5 days Newbie friendliness 42/100

deepset-ai/haystack#12777 · 2 comments · Opened Sep 16, 2026
CSVDocumentCleaner silently removes nonempty NA and NULL strings Open

deepset-ai/haystack#12771 · 1 assignee · Opened Sep 16, 2026
JsonSchemaValidator raises IndexError for an empty message list Open

deepset-ai/haystack#12765 · 1 assignee · Opened Sep 16, 2026
JsonSchemaValidator rejects valid empty schemas and ignores empty runtime overrides Open

deepset-ai/haystack#12763 · 1 assignee · Opened Sep 16, 2026
RecursiveDocumentSplitter split_idx_start is off by the trailing separator with split_unit="word" and split_overlap > 0 Open

deepset-ai/haystack#12756 · 1 assignee · Opened Sep 15, 2026
InMemoryDocumentStore unique metadata count fails for list and dict values Open

deepset-ai/haystack#12746 · 1 comment · 1 assignee · Opened Sep 14, 2026
LLMDocumentContentExtractor treats a chat generator's own "error" field as a failed call Open

deepset-ai/haystack#12734 · 1 assignee · Opened Sep 13, 2026
Optional Vynly Integration for Visual Content Generation Open

P3

Difficulty 5/5 Over a week Newbie friendliness 25/100

deepset-ai/haystack#12728 · Opened Sep 13, 2026
Bug: MarkdownHeaderSplitter shares parent_headers list across secondary chunks when keep_headers=False Open

P2

deepset-ai/haystack#12691 · 1 comment · 1 assignee · Opened Sep 10, 2026
DocumentSplitter emits redundant overlap-only trailing chunks for character-based split modes (token mode was fixed in #12661) Open

deepset-ai/haystack#12686 · 1 assignee · Opened Sep 10, 2026
OpenAIResponsesChatGenerator discards annotations Open

P1

deepset-ai/haystack#12682 · 1 comment · 1 assignee · Opened Sep 9, 2026
_merge_super_component_pipelines docstring contradicts the function signature and implementation Open

deepset-ai/haystack#12656 · 1 comment · 1 assignee · Opened Sep 7, 2026
DocumentJoiner positional weights follow Variadic arrival order, not connect() order Open

deepset-ai/haystack#12646 · 2 comments · 1 assignee · Opened Sep 7, 2026
File-level and inline comments in breakpoint.py say snapshot saving is 'enabled by default' (it is disabled) Open

P2

Difficulty 2/5 1-3 hours Newbie friendliness 88/100

deepset-ai/haystack#12643 · 1 comment · Opened Sep 7, 2026
OutputAdapter silently coerces string output to another type when output_type is a Union including str. Open

P2

deepset-ai/haystack#12625 · 1 comment · 1 assignee · Opened Sep 5, 2026
Agent tool scheduler picks write-write winner by batch order, not call order, contradicting its own docstring and #11690 Open

deepset-ai/haystack#12621 · 3 comments · 1 assignee · Opened Sep 5, 2026
ConditionalRouter silently coerces string output to another type when output_type is `str | None` Open

deepset-ai/haystack#12617 · 1 assignee · Opened Sep 5, 2026
ChatMessage.from_openai_dict_format silently collapses OpenAI's 'developer' role into 'system' Open

P2

deepset-ai/haystack#12604 · 1 comment · 1 assignee · Opened Sep 4, 2026
[HTTPXodus] Consider migrating from `httpx` to `httpx2` (the actively maintained fork) Open

P3

deepset-ai/haystack#12593 · 1 assignee · Opened Sep 4, 2026
4 of 5 Document*Evaluators crash with ZeroDivisionError on empty input Open

P3

Difficulty 3/5 1-2 days Newbie friendliness 84/100

deepset-ai/haystack#12543 · 3 comments · Opened Sep 1, 2026
CacheChecker.run issues one filter_documents call per item (N+1 query pattern) Open

P2

Difficulty 4/5 3-5 days Newbie friendliness 20/100

deepset-ai/haystack#12535 · 4 comments · Opened Aug 31, 2026
Feature: optional EverOS long-term memory integration Open

P3

deepset-ai/haystack#12499 · 2 comments · 1 assignee · Opened Aug 28, 2026
OpenAI and Azure Responses Chat Generators fail to deserialize ComponentTool Open

deepset-ai/haystack#12484 · 1 comment · 1 assignee · Opened Aug 27, 2026
DocumentToImageContent raises ValueError on bad inputs, blocking LLMDocumentContentExtractor partial-success Open

P2

deepset-ai/haystack#12447 · 1 comment · 1 assignee · Opened Aug 24, 2026
ChatMessage.from_openai_dict_format wraps list-of-parts content in TextContent, breaking msg.text contract Open

deepset-ai/haystack#12445 · 1 assignee · Opened Aug 24, 2026
fix: serialize_type / deserialize_type do not support typing.Annotated Open

P2

deepset-ai/haystack#12338 · 1 comment · 1 assignee · Opened Aug 13, 2026
PythonCodeSplitter with strip_docstrings=True emits invalid Python for docstring-only function/method/class bodies Open

deepset-ai/haystack#12330 · 2 comments · 1 assignee · Opened Aug 13, 2026
Add `FilterBuilder` to make Haystack metadata filters easier to write and read Open

P3

Difficulty 5/5 Over a week Newbie friendliness 48/100

deepset-ai/haystack#12157 · 3 comments · Opened Jul 26, 2026
Allow Human-in-the-Loop to review fully materialized tool calls Open

P2

Difficulty 5/5 Over a week Newbie friendliness 35/100

deepset-ai/haystack#12060 · 4 comments · Opened Jul 18, 2026
RFC: Retrieval Diagnostics API for RAG Pipelines Open

P3

Difficulty 5/5 Over a week Newbie friendliness 35/100

deepset-ai/haystack#11867 · 2 comments · Opened Jul 3, 2026
feat: Pipeline run recording & deterministic replay - turn every run into a shareable, diffable, testable artifact Open

P3

Difficulty 5/5 Over a week Newbie friendliness 25/100

deepset-ai/haystack#11836 · 9 comments · Opened Jun 30, 2026
Support for websocket mode in OpenAIResponsesChatGenerator Open

P2

Difficulty 5/5 Over a week Newbie friendliness 20/100

deepset-ai/haystack#11799 · 3 comments · Opened Jun 29, 2026
Cache tool results inside Agent loops (avoid duplicate identical tool calls) Open

P3

Difficulty 5/5 Over a week Newbie friendliness 38/100

deepset-ai/haystack#11588 · 3 comments · Opened Jun 11, 2026
Unify request-scoped state in a single PipelineRunContext (extends #11366) Open

P3

Difficulty 5/5 Over a week Newbie friendliness 35/100

deepset-ai/haystack#11580 · 3 comments · Opened Jun 10, 2026
Native `map` / fan-out component for iterating a sub-pipeline over a list with aggregation Open

P3

Difficulty 5/5 Over a week Newbie friendliness 20/100

deepset-ai/haystack#11521 · 2 comments · Opened Jun 5, 2026
feat: Add EmbeddingCache component to avoid re-embedding identical text Open

P3

Difficulty 5/5 Over a week Newbie friendliness 25/100

deepset-ai/haystack#11476 · 3 comments · Opened Jun 2, 2026
feat: support token-based budget in LostInTheMiddleRanker Open

P3

Difficulty 3/5 1-2 days Newbie friendliness 72/100

deepset-ai/haystack#11351 · 3 comments · Opened May 20, 2026
RFC: Structured Evaluator Uncertainty and Error Semantics Open

P3

Difficulty 5/5 Over a week Newbie friendliness 25/100

deepset-ai/haystack#11332 · 5 comments · Opened May 18, 2026
Proposal: Transaction Protocol for idempotent, auditable agent pipelines Open

P3

Difficulty 5/5 Over a week Newbie friendliness 20/100

deepset-ai/haystack#11266 · 7 comments · Opened May 6, 2026
docs: Update Ragas docs Open

P2 type:documentation

Difficulty 3/5 1-2 days Newbie friendliness 45/100

deepset-ai/haystack#11178 · 2 comments · Opened Apr 23, 2026
Add optional retrieval confidence to retriever outputs Open

P3

Difficulty 5/5 Over a week Newbie friendliness 35/100

deepset-ai/haystack#11177 · Opened Apr 23, 2026

deepset-ai/haystack

Get new issues in your inbox