v0.2.3 - PyPI Discoverability & Documentation Standards

This release focuses on optimizing DagPipe's presence on PyPI and standardizing the documentation syntax to ensure a professional, technical presentation.

Enhancements
PyPI Discoverability: Optimized the project description and added specialized classifiers (Scientific/Engineering :: Artificial Intelligence) to improve search rankings and categorization on PyPI.
Documentation Refactor: Systematically replaced informal punctuation (em dashes) with standard structural delimiters (commas, colons, and semicolons) throughout the

README.md
for better readability.
Syncing Test Metrics: Updated documented test coverage from 108 to 132 passed tests to accurately reflect the current state of the battle-hardened core.

Maintenance
Synchronized version strings to 0.2.2 across the core package,

llms.txt
(agent primer),

server.json
, and the library source code (

src/dagpipe/
init
.py
).
Verified that all 132 tests pass with zero regressions.
Installation

bash
pip install dagpipe-core==0.2.2
View this release on PyPI

v0.2.0 — Smart Router + Model Registry

What's New

Smart Model Router

Automatically routes tasks to the cheapest model that can handle them. Simple tasks go to free-tier models. Hard tasks escalate. Rate limit failures trigger automatic fallback.

Live Model Registry

Self-maintaining database of free-tier model availability and pricing. Refreshes every 24 hours automatically.

Semantic Output Assertions

Nodes can declare logical rules on their output. Structurally valid but semantically wrong data is caught and forced to self-correct.

Context Isolation

Nodes only receive outputs from their declared dependencies. Untrusted data cannot bleed across pipeline branches.

Dead Letter Queue

Full failure context saved to disk when a node exhausts all retries. No failure information is ever lost.

Pluggable Checkpoint Backends

Swap filesystem storage for Redis, S3, or any custom backend via the CheckpointStorage protocol.

Backwards Compatibility

All v0.1.x pipelines run without changes. Every new feature is opt-in.

Stats

108 tests · 0 regressions · Python 3.12 + 3.13

v0.1.0 — Core Library.

DagPipe v0.1.0 — Core Library

First public release. Phase 1 complete.

What's included

• dag.py — Generic DAG orchestrator with topological sort
• checkpoints.py — JSON crash recovery
• constrained.py — Pydantic-validated LLM output with retry
• router.py — Complexity-based free-tier model routing

Install

pip install dagpipe-core

Stats

37 tests passing · 0 paid dependencies · Python 3.12+