2.5.0

v2.5.0 (2026-01-19)

Bug Fixes

• Load step settings from prefixed environment variables in Argo backend (#194, d1b6570)

• Pin pathspec DVC test error (#209, bb3304f)

Chores

• deps: Bump reuse from 5.0.2 to 6.2.0 (#201, cc829a3)

• deps: Bump ruff from 0.14.4 to 0.14.10 (#204, da812bb)

• deps: Update pre-commit requirement from ==4.1.* to >=4.1,<4.6 (#189, 9e69d53)

• deps: Update pytest requirement from ==8.4.* to >=8.4,<9.1 (#187, 47408ba)

Continuous Integration

• Run docker img on pull_request (#139, 64aaaaf)

Features

• Add Decagon Knowledge Base push step (#208, 33a57b6)

• Add environment variable management CLI command (#192, 56e2594)

• Loading argo config from yaml file (#197, a068d72)

• Update env command to prefer current environment values in .env… (#193, 1ee3030)

---

Detailed Changes: 2.4.0...2.5.0

2.4.0

v2.4.0 (2025-11-20)

Bug Fixes

• Limit token count based on offset in HF tokenizer (#185, 6e9121c)

• Rework backend CLI (#160, b783fcd)

• Setting token length correctly in splitter metadata (#186, bdf30a5)

Chores

• deps: Bump pymilvus from 2.5.14 to 2.6.3 (#173, 4e854d6)

• deps: Bump ruff from 0.13.2 to 0.14.4 (#181, 695f5a4)

• deps: Bump tiktoken from 0.11.0 to 0.12.0 (#174, a01ab59)

• deps: Bump tqdm from 4.66.5 to 4.67.1 (#172, 212a6e2)

Features

• Adding TruncatedEmbeddingStep (2) and document metadata to Qdrant (#183, a92d004)

• Splitter tracks token and char length, embedding step logs the length statistics (#167, 4a276d1)

---

Detailed Changes: 2.3.0...2.4.0

2.3.0

v2.3.0 (2025-11-03)

Chores

• deps: Bump mistletoe from 1.4.0 to 1.5.0 (#164, 9035f5f)

Features

• Add sftp step (#151, 5305b98)

---

Detailed Changes: 2.2.1...2.3.0

2.2.1

v2.2.1 (2025-10-29)

Bug Fixes

• Hf tokenizers decoding skips special tokens (#170, 0c2f6e8)

• Splitter step fails if it has no results (#159, 802bf43)

Chores

• deps: Bump dvc from 3.53.1 to 3.63.0 (#162, 506c2de)

• deps: Bump prometheus-client from 0.21.1 to 0.23.1 (#165, 7d7fd92)

• deps: Bump tiktoken from 0.7.0 to 0.11.0 (#153, de28bb3)

• deps: Update lxml requirement from ==5.2.* to >=5.2,<6.1 (#152, 762c70e)

• deps: Update pytest-cov requirement from ==4.* to >=4,<8 (#155, 50f422d)

---

Detailed Changes: 2.2.0...2.2.1

2.2.0

v2.2.0 (2025-10-02)

Bug Fixes

• Clean up wurzel folders (#137, 1ee4fb8)

• Docs & add workflows (#122, 207dedc)

• Enforcing token minimum on all recursive depths (#121, ceb201a)

• Fixing deprecated warning in MdDataContract (#130, 236d2cd)

• Readme #119 (#136, 299fb8c)

• Rm error msg on init of dvc (#113, 0f3d648)

• Speed up wurzel cli (#135, 8d4acf4)

• Uncaught_exception_logging (#140, 86ec7ca)

• Unpin apt deps in Dockerimg (#132, fdc3bd9)

• Yaml-metadata for Markdown files following the Jekyll standard (fixes #101) (#120, 28217ee)

Chores

• deps: Bump mdformat from 0.7.17 to 0.7.22 (#148, 783f438)

• deps: Bump pandas from 2.2.2 to 2.3.3 (#147, abf5b66)

• deps: Bump qdrant-client from 1.10.1 to 1.15.1 (#144, 86f99c9)

• deps: Bump ruff from 0.9.10 to 0.13.2 (#145, d5f457d)

Continuous Integration

• Add semantic commit ci (#124, 6e46a8f)

• One pr for every dep (#143, d2800ef)

• Run tests with min install (#127, 91fa0ec)

Documentation

• Update and improve documentation structure (#126, 3784e35)

Features

• Adding abstraction for sentence splitting (#117, 9089399)

• Adding tokenizer abstraction to handle OpenAI's tiktoken and HF tokenizers (#114, 54e2cd8)

• Adding wtpsplit's SaT sentence splitter (#125, ad90ee8)

• Load SpaCy model via CLI at runtime (#138, 252df68)

---

Detailed Changes: 2.1.3...2.2.0

2.1.3

v2.1.3 (2025-08-11)

Bug Fixes

• For table splitter to avoid cutting off content (#96, 72a1537)

• Splitter docs (#112, c3c270b)

Ref

• Call of MarkdownTableSplitter (#96, 72a1537)

---

Detailed Changes: 2.1.2...2.1.3

2.1.2

v2.1.2 (2025-08-05)

Bug Fixes

• Secretstr (#111, f12d3a3)

---

Detailed Changes: 2.1.1...2.1.2

2.1.1

v2.1.1 (2025-07-31)

Bug Fixes

• Pypi release (#108, 73e5c6f)

---

Detailed Changes: 2.1.0...2.1.1

2.1.0

v2.1.0 (2025-07-31)

Bug Fixes

• Convert html to md table (#98, ac4a8cb)

• Convert html to md table (#95, bcdbc80)

• Tags on ghr (#98, ac4a8cb)

Chores

• Add Sample pipeline in README (#100, 0c0ff2b)

Continuous Integration

• Add new ruff rule #104 (#105, 1b46cc5)

Features

• Adding logging of discarded text in splitter step (#99, 259a1ea)

• Docker file ref & use uv (#98, ac4a8cb)

• Update docker file (#98, ac4a8cb)

---

Detailed Changes: 2.0.0...2.1.0

2.0.0

v2.0.0 (2025-07-18)

Features

• Argo workflow backend (1714f3f)

---

Detailed Changes: 1.3.0...2.0.0