A personal radar for one learner (Python-first, Azure + cloud-agnostic, Java/Spring AI alongside; goal Staff/AI architect).
Rings answer "what should I do with this in the next 24 weeks?", not "is it good?".
Weekly agent (Sunday ~07:00 IST, see AGENTS.md §6.2): re-verifies versions and status of every blip against primary sources
(release notes, PyPI, official docs), moves rings when evidence changes, updates the "as of" date, and adds a dated line to the changelog below.
Daily agent flags releases affecting any blip in its digest; it does not edit the radar, it only proposes a move in the digest.
Ring-move rules:
Assess → Trial: stable release (not preview) + a second independent production report + fits a roadmap task.
Trial → Adopt: used in a lab or the capstone without regret, and a clear "when not to use" is known.
Any → Hold: vendor maintenance-mode announcement, retirement date, or repeated evidence it underperforms simpler baselines.
Blip format: name, ring, quadrant, 1–2 line rationale, as of date, source link. Prefer primary sources. Version claims must say "as of".
Human override: if you built something and disagree with a ring, edit the row and add a changelog line with your reason.
Initial radar published (56 blips). Versions checked against PyPI/official docs on this date: LangGraph 1.2.x, Pydantic AI 2.x, DSPy 3.4, Microsoft Agent Framework 1.x, Google ADK 2.x, MCP Python SDK 2.x, Zensical 0.0.x.