Repository navigation
Conversation
…breakage workflows
…ort Dispatcher.yml
a0e4568 to
0baf50a
Compare
Phase A1 — Declarative Dispatch Engine & Unit TestsCommit: What was done:
|
Phase A2 — Reusable Dispatch Workflow & Workflow HardeningCommit: What was done:
|
Phase B — Release Tooling IntegrationCommit: What was done:
|
Phase C — Documentation & Catalog EntryCommit: What was done:
|
Critical review of #148Conventions applied: ct-conventions v0.11.0, ct-agents v0.4.0. Reviewed at The direction is right: one PR entry point, declarative 🔴 BlockersB1. An unrelated label silently invalidates
Adding any cosmetic label after CI turns the PR "not ready", and the only way out is to re-run everything. Possible fixes:
B2. Adding one label re-runs every labelled suite, GPU included.
B3. 🟠 MajorM1. Label and default knowledge is duplicated in five places, and they already disagree.
As a result, the configurable labels in M2.
M3. M4. The "< 2 s" claim is misleading. M5. Multi-line values corrupt M6. 🟡 Minor
🧪 Tests
VerdictRequest changes. Fix B1–B3 and M1–M2 before the pilot on OptimalControlProblems.jl. Otherwise the pilot will reproduce exactly the problems #144 is meant to fix (spurious "not ready", lost runner time). B2 needs a human design decision (full re-run vs. per-label aggregation in |
Fixes applied addressing the critical reviewCommit: 1. Blockers resolved:
2. Major issues resolved:
Check results:
|
|
Closing without merging. The single The actual problem (runs hidden on the PR page and ~30 skipped jobs per label) is tracked in #149, with a different approach (label → |
Refs #144
Introduce a centralized, unified label dispatcher and PR gate (
dispatch.yml) in CTActions, powered by declarative.github/ci.tomlrepository configuration.