Haystack
Cooling · Discussion moved -32.2% and downloads moved -24.7% — both receding. (as of 2026·08·12)
Narrow: 83 distinct voices behind 248 mentions.
Distinct voices per mention. 1.00 means every mention came from a different person; a low number means a handful of people are driving the volume. Authors are counted once per day, so someone who posts across several days counts more than once over the whole window. The figure can overstate breadth, never understate it.
A wave of praise on a forum and a rapidly filling issue tracker both show up as “mentions up” until you see where they came from.
Opinion is genuinely divided. The average sentiment sits near zero because the two sides cancel out, not because people are lukewarm.
SRC: 702 mentions in window · net sentiment 0.13 · 3 adoption series · derived aggregates only
Interactive chart. Arrow keys step through the series, Home and End jump to its ends, and Escape closes the readout.
| Period | mentions |
|---|---|
| 2026-06-11 | 1 |
| 2026-07-05 | 1 |
| 2026-07-09 | 1 |
| 2026-07-10 | 2 |
| 2026-07-11 | 1 |
| 2026-07-12 | 1 |
| 2026-07-13 | 24 |
| 2026-07-14 | 26 |
| 2026-07-15 | 31 |
| 2026-07-16 | 26 |
| 2026-07-17 | 48 |
| 2026-07-18 | 7 |
| 2026-07-19 | 11 |
| 2026-07-20 | 40 |
| 2026-07-21 | 42 |
| 2026-07-22 | 19 |
| 2026-07-23 | 31 |
| 2026-07-24 | 25 |
| 2026-07-25 | 15 |
| 2026-07-26 | 11 |
| 2026-07-27 | 27 |
| 2026-07-28 | 37 |
| 2026-07-29 | 27 |
| 2026-07-30 | 21 |
| 2026-07-31 | 43 |
| 2026-08-01 | 6 |
| 2026-08-02 | 10 |
| 2026-08-03 | 21 |
| 2026-08-04 | 41 |
| 2026-08-05 | 29 |
| 2026-08-07 | 1 |
| 2026-08-08 | 1 |
| 2026-08-09 | 1 |
| 2026-08-11 | 32 |
| 2026-08-12 | 42 |
Interactive chart. Arrow keys step through the series, Home and End jump to its ends, and Escape closes the readout.
| Period | net sentiment |
|---|---|
| 2026-06-11 | 0.97 |
| 2026-07-05 | 0.98 |
| 2026-07-09 | −0.61 |
| 2026-07-10 | −0.12 |
| 2026-07-11 | 0.98 |
| 2026-07-12 | −0.73 |
| 2026-07-13 | 0.12 |
| 2026-07-14 | 0.12 |
| 2026-07-15 | 0.14 |
| 2026-07-16 | 0.17 |
| 2026-07-17 | 0.13 |
| 2026-07-18 | 0.40 |
| 2026-07-19 | 0.21 |
| 2026-07-20 | 0.07 |
| 2026-07-21 | 0.23 |
| 2026-07-22 | −0.02 |
| 2026-07-23 | 0.27 |
| 2026-07-24 | 0.41 |
| 2026-07-25 | 0.58 |
| 2026-07-26 | 0.64 |
| 2026-07-27 | −0.03 |
| 2026-07-28 | 0.12 |
| 2026-07-29 | 0.21 |
| 2026-07-30 | 0.12 |
| 2026-07-31 | 0.29 |
| 2026-08-01 | 0.37 |
| 2026-08-02 | 0.32 |
| 2026-08-03 | 0.17 |
| 2026-08-04 | 0.20 |
| 2026-08-05 | 0.20 |
| 2026-08-07 | 0.92 |
| 2026-08-08 | 0.81 |
| 2026-08-09 | 0.94 |
| 2026-08-11 | −0.11 |
| 2026-08-12 | 0.13 |
Interactive chart. Arrow keys step through the series, Home and End jump to its ends, and Escape closes the readout.
| Period | GitHub releases / day |
|---|---|
| 2026-06-02 | 1 |
| 2026-06-03 | 1 |
| 2026-06-09 | 2 |
| 2026-06-17 | 1 |
| 2026-06-18 | 1 |
| 2026-07-07 | 2 |
| 2026-07-08 | 1 |
| 2026-07-20 | 2 |
Interactive chart. Arrow keys step through the series, Home and End jump to its ends, and Escape closes the readout.
| Period | GitHub stars (total) |
|---|---|
| 2026-08-06 | 26.1k |
| 2026-08-07 | 26.1k |
| 2026-08-08 | 26.1k |
| 2026-08-09 | 26.2k |
| 2026-08-10 | 26.2k |
| 2026-08-11 | 26.2k |
| 2026-08-12 | 26.2k |
Interactive chart. Arrow keys step through the series, Home and End jump to its ends, and Escape closes the readout.
| Period | PyPI downloads / day |
|---|---|
| 2026-05-15 | 41.3k |
| 2026-05-16 | 26.2k |
| 2026-05-17 | 27.1k |
| 2026-05-18 | 36.0k |
| 2026-05-19 | 37.3k |
| 2026-05-20 | 34.7k |
| 2026-05-21 | 35.8k |
| 2026-05-22 | 32.0k |
| 2026-05-23 | 23.7k |
| 2026-05-24 | 26.3k |
| 2026-05-25 | 28.6k |
| 2026-05-26 | 33.3k |
| 2026-05-27 | 35.3k |
| 2026-05-28 | 33.9k |
| 2026-05-29 | 32.4k |
| 2026-05-30 | 27.0k |
| 2026-05-31 | 22.7k |
| 2026-06-01 | 29.2k |
| 2026-06-02 | 31.2k |
| 2026-06-03 | 30.4k |
| 2026-06-04 | 31.3k |
| 2026-06-05 | 28.7k |
| 2026-06-06 | 20.7k |
| 2026-06-07 | 21.2k |
| 2026-06-08 | 29.6k |
| 2026-06-09 | 30.3k |
| 2026-06-10 | 32.1k |
| 2026-06-11 | 33.9k |
| 2026-06-12 | 30.8k |
| 2026-06-13 | 24.5k |
| 2026-06-14 | 27.8k |
| 2026-06-15 | 38.1k |
| 2026-06-16 | 36.4k |
| 2026-06-17 | 36.6k |
| 2026-06-18 | 36.5k |
| 2026-06-19 | 44.1k |
| 2026-06-20 | 32.3k |
| 2026-06-21 | 30.5k |
| 2026-06-22 | 41.4k |
| 2026-06-23 | 40.4k |
| 2026-06-24 | 41.0k |
| 2026-06-25 | 38.9k |
| 2026-06-26 | 34.0k |
| 2026-06-27 | 25.9k |
| 2026-06-28 | 27.1k |
| 2026-06-29 | 35.5k |
| 2026-06-30 | 37.3k |
| 2026-07-01 | 38.7k |
| 2026-07-02 | 39.7k |
| 2026-07-03 | 34.8k |
| 2026-07-04 | 25.8k |
| 2026-07-05 | 24.9k |
| 2026-07-06 | 34.0k |
| 2026-07-07 | 37.8k |
| 2026-07-08 | 37.8k |
| 2026-07-09 | 34.8k |
| 2026-07-10 | 35.8k |
| 2026-07-11 | 25.8k |
| 2026-07-12 | 24.3k |
| 2026-07-13 | 33.8k |
| 2026-07-14 | 36.2k |
| 2026-07-15 | 32.3k |
| 2026-07-16 | 42.0k |
| 2026-07-17 | 44.9k |
| 2026-07-18 | 23.7k |
| 2026-07-19 | 25.2k |
| 2026-07-20 | 38.4k |
| 2026-07-21 | 44.3k |
| 2026-07-22 | 44.2k |
| 2026-07-23 | 49.1k |
| 2026-07-24 | 30.6k |
| 2026-07-25 | 19.6k |
| 2026-07-26 | 20.1k |
| 2026-07-27 | 31.1k |
| 2026-07-28 | 29.9k |
| 2026-07-29 | 30.9k |
| 2026-07-30 | 31.4k |
| 2026-07-31 | 30.5k |
| 2026-08-01 | 22.9k |
| 2026-08-02 | 22.6k |
| 2026-08-03 | 29.7k |
| 2026-08-04 | 31.8k |
| 2026-08-05 | 30.5k |
| 2026-08-06 | 26.0k |
| 2026-08-07 | 27.7k |
| 2026-08-08 | 19.8k |
| 2026-08-09 | 19.4k |
| 2026-08-10 | 32.5k |
| 2026-08-11 | 32.0k |
| What was said | Momentum |
|---|---|
| Agent loops to max_agent_steps when a reasoning model returns no text because it hits max_output_tokens. | +0% |
| LinkContentFetcher fails to properly rotate User-Agent headers independently across concurrent multi-URL fetch retries. | +0% |
| RecursiveDocumentSplitter with overlap produces chunks containing text not in the source document. | +0% |
| Pipeline connection type conversion is non-deterministic across processes due to unguarded set.pop() usage. | +0% |
| The _get_conversion_strategy function uses non-deterministic set.pop() on enum members when falling back, causing different behavior across processes due to ASLR. | +0% |
| The _apply_overlap function is called at every recursion level of _chunk_text, causing inner-level chunks to get overlap prepended multiple times and producing chunks that are not substrings of the source. | +0% |
| The serialize_type and deserialize_type functions cannot round-trip typing.Literal type hints, causing pipelines using Literal types to crash on load. | +0% |
| Human-in-the-loop rejection turns into an approved tool call when only some decisions carry a tool_call_id. | +0% |
| MockChatGenerator.response_fn should receive tools passed to run so dynamic mocks can produce replies that depend on available tools and their schemas. | +0% |
| Document.__eq__ ignores colliding metadata keys during equality comparison. | +0% |
| InMemoryDocumentStore retains zero-frequency vocabulary entries after document deletion, affecting BM25 scoring. | +0% |
| Add runtime verification for MCP pipelines in Haystack to prevent injection and path traversal attacks. | +0% |
| FileSystemToolResultStore.read() allows reading files outside the configured store root. | +0% |
| FilterPolicy.MERGE silently drops init filters in InMemoryBM25Retriever and InMemoryEmbeddingRetriever. | +0% |
| Resuming a pipeline snapshot after the first loop visit fails with PipelineComponentsBlockedError. | +0% |
| Integrate CCS runtime verification as a pipeline component in Haystack's tool execution chain. | +0% |
| Add FilterBuilder to make Haystack metadata filters easier to write and read. | +0% |
| Exact-match metadata filters silently miss equivalent ISO timestamps in different formats (Z vs +00:00). | +0% |
| FallbackChatGenerator repeatedly calls warm_up() on every run invocation instead of once. | +0% |
| Toolset.add() silently drops tools from lazy Toolsets before parent warm-up completes. | +0% |