quantmodel

← Results index

Validate SPYST parser on history and add a manual-review fallback

2026-08-15T21:44:36Z Research completed Ran the parser over historical emails with known interpretations to measure state/allocation accuracy, success count and manual-review rate; ambiguous, out-of-range or layout-changed mail is flagged for review and corrections are retained as fixtures so a publisher layout change fails loudly. Commit 5263cc97 added research/spyst_signals/validate.py (437 lines), chart correction/signal fixtures and 344-line validation tests.
qm-xz0
Reconstructed — uncertain: some details inferred from git history/artifacts and marked below. Unsupported details omitted.

Objective

Complete pipeline task (#811) (commit 5263cc97 — qm-xz0.3: Complete pipeline task (#811)).

Approach

[reconstructed — uncertain] Reconstructed from git commit 5263cc97 and changed paths. No dedicated findings markdown found.

Files / code / data changed

 research/spyst_signals/allocation.py               |  27 +-
 .../data/spyst_chart_corrections.json              |  43 ++
 .../spyst_signals/data/spyst_chart_signals.json    |  62 +++
 .../data/spyst_validation_report.json              |  50 +++
 research/spyst_signals/extract.py                  |  65 ++-

Results

[reconstructed — uncertain] Commit message: qm-xz0.3: Complete pipeline task (#811). See commit diff and referenced findings for metrics.

Conclusions

Commit 5263cc97 merged via PR; outcome inferred from commit success and subsequent build.

Problems / follow-ups

Links

Source artifact: research/results/qm-xz0.3.md