Show each review round's failure-class chains, churn and cost in status

Aperta
  • Qualcosa non va?
#458 0 commenti 0 reazioni 0 assegnatari Vedi su GitHub

I maintainer di solito rispondono entro 1 giorno

Nessuno ha ancora preso questa issue.

Valutazione

Difficoltà
4/5
Tempo stimato
3-5 giorni
Idoneità per principianti
35/100
Tipo di issue
Funzionalità
Chiarezza
Abbastanza chiara
Stato di attività
Attiva
Stack tecnologico
python
Ambito
cli, observability

Direzione di ricerca

Start with the status output and the per-round run-log record, and read execution.json to see where the Claude-reported cost is stored. The churn figures must come from the controller's own git reads of each FIX commit (--shortstat, cumulative since the first FIX), not from agent output. This issue depends on #452 and #453, which supply chain ids and lengths, so check those first. Done means a fixture history renders each round's findings, chains, churn and cost, showing "unknown" or "not reported" where data is absent.

Scritto dal modello di indicizzazione a partire dal testo della issue.

Descrizione

enhancement

Problem

Nothing in status or the run log shows whether a REVIEW/FIX loop is converging. An operator watching Lychee-Technology/forma#801 saw finding counts of 1, 1, 1… and could not tell, without reading every review comment, that each round was the same unresolved problem and that each FIX was adding hundreds of lines to one module.

Evidence

From #451:

  • Churn was large but invisible. Rounds 3–11 of Lychee-Technology/forma#801 added +2,128 / −473 lines, about 95% of them in the scanner and lexer files.
  • Cost is only partly recorded. The Claude-reported cost is in execution.json ($24.42 for the run), but OpenCode's provider_summary holds no cost at all, so the total cost of the incident is unknown.

Desired outcome

status, and a run-log record per review round, show:

  • the finding count;
  • each finding's chain id and chain length (from #452);
  • the FIX commit's --shortstat and the cumulative change since the first FIX, computed by the controller from git, never taken from agent claims;
  • the provider-reported cost where there is one, and "not reported" where there is not.

Acceptance criteria

  • status renders these from a fixture history. A history without links or costs renders "unknown" or "not reported", never zero.
  • The churn numbers come from the controller's own git reads of the reported commits.
  • No new GitHub read or write.

Trade-offs and unknowns

Whether OpenCode can report cost at all in this mode is unknown; if it cannot, the field stays "not reported".

Depends on #452 and #453.

Lingua principale
Python
Stelle
0
Fork
0
Merge medio
5h 42m
PR unite (30g)
110

Preparare l'ambiente

Questo progetto non fornisce container di sviluppo, Dockerfile né guida per i contributori, quindi l'ambiente è a tuo carico: parti dal suo README e consulta la nostra guida al primo contributo per i passaggi generali.

Come iniziare

  1. Leggi tutta la issue e poi la guida ai contributi del progetto.
  2. Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
  3. Fai un fork del repository e lavora su un branch.
  4. Apri una pull request che faccia riferimento al numero della issue.

Altre issue di iceboundrock/AutoForge

Tutte le issue di iceboundrock/AutoForge

Issue simili

Altre issue su Python

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.