Skip to content

feat: cortex sync-runs — run status from a campaign digest (hpc-campaign p5b) #82

Closed

Description

@Jammy2211

Overview

Phase 5b: cortex sync-runs flips run lines from a campaign digest under decision 62 (ruling F3); result/lesson stay human. Branch feature/hpc-campaign-p5b-cortex. Tier judge.

Heart RED at the door (2026-10-30): release validation FAILED (stage integrate); PyAutoHeart/workflows/Release Integrate failure; PyAutoNerves test_fitsable.py::test__output_to_fits red; Unit-test timing 1 regression (>3× baseline), 14 slow; autogalaxy red; Release validation NOT release_ready (v2026.10.30.1.dev85101); autolens_profiling red; Worktree drift 1 dirty. Human override for Heart RED (development only) authorized live by the human for hpc-campaign phase 5 ("Override + approve"). Merge needs green CI and a separate /prm; no release.

Original Prompt

Click to expand

Issued: 2026-10-30
Epic: hpc-campaign
Phase: 5b
Blocked-by: phases 2–4 (COMPLETE 2026-10-30); ruling F3 (yes)
Consequence: judge
Autonomy: human-required

HPC campaign phase 5b — cortex sync-runs from a campaign digest

Original request (human, 2026-10-30): "spread across grid bases, prm on those and then we do next phase 5"; plan
approved with "Override + approve" (5a assistant, 5b Cortex, 5c Brain in parallel; 5d pipeline adoption deferred
until DR1 v1.0 ships).

Lane prompt: Workspace/Science/scratch/pyautolens_ii_campaign_results_scoping_2026-10-28/PROMPT_hpc_campaign.md §Phase 5.
@PyAutoCortex: cortex sync-runs <project> --digest <file> flips open→running and calls done with finished/failed counts (never result/lesson); projects.yaml campaign_file + digest fields; decision 62 text per ruling F3; REFERENCE boundary paragraph (status is a cluster fact; result/lesson stay human). Tests + cortex check.

Activity

  1. Jammy2211 commented on Oct 30, 2026

    @Jammy2211
    ContributorAuthor

    Status 2026-10-30 (session wrap-up)

    Lane built locally: 1bc7ef7 on feature/hpc-campaign-p5b-cortex (worktree ~/Code/PyAutoLabs-wt/hpc-campaign-p5b-cortex), 87 tests, cortex check OK — not pushed, no PR. Decision number used: 64 (62/63 taken).

    Independent review: FINDINGS — fix before PR:

    1. HIGH _tasks_of (~984) + all-terminal branch (~1048): pending array slices appear in sacct as one compound row (job_id: 345242_[0-8]); they never match, and nothing checks every index of the run ident is covered, so a run with pending tasks is marked done and its line removed (irreversible). Repro: run 3002_[0-3] with rows 3002_0,3002_1 completed + 3002_[2-3] pending → done — 2 completed. Fix: expand compound job_id/key ranges; refuse done unless every ident index has a terminal task; a pending stub blocks a bare-id run.
    2. MEDIUM sync_runs loop (~1036) writes per run; a later CortexError leaves earlier runs changed with empty stdout; _write (~553) writes before validating. Fix: compute all edits, write once.
    3. LOW open→running _prepend_entry (~1060) skips add_entry's head-date check (newest-first).

    Real-data dry run (temp copy): 58 euclid_dr1 runs would change — re-check after fix 1.

  2. Jammy2211 commented on Oct 30, 2026

    @Jammy2211
    ContributorAuthor

    Shipped

    Merged via #83 — cortex sync-runs (decision 64). Review findings fixed (d590eac). Applying it to the live euclid_dr1 ledger (58 runs would close) is a separate human ask.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions