Hacktoberfest 2026：メンテナが10月に向けて印を付けた、オープンで初心者向けの issue。 Hacktoberfest の issue を見る

[Phase 16] S11 — Deploy/serving generator contract

オープン
  • 何か問題がありますか？
#897 コメント 0 件 リアクション 0 件 担当者 0 名 GitHub で見る

まだ誰も着手していません。

評価

難易度
4/5
見積もり時間
3〜5日
初心者へのやさしさ
30/100
issue の種類
バグ
明瞭さ
おおむね明確
活発さ
活発
技術スタック
docker-compose, dockerfile, python
領域
devops

調査の方向性

Start with the S11 section of docs/roadmap/phase-16-trust-surface-hardening.md (lines 613-634) and the deploy generator's generate_deploy_config(), then compare the manual's TGI and HF Endpoints sections against what the generator emits. The issue spans four linked units and a bilingual EN/TR manual, so begin with the golden-test harness the Plan describes. Done when the exit criteria pass: manual commands match real output, the --output existing-dir case returns success=False, and the compose file has a non-latest image and read-only /data mount.

索引モデルが issue の本文から書いたものです。

説明

phase: 16 source: roadmap

Roadmap · Phase 16, step S11 · size M · 4 unit(s) · planned in docs/roadmap/phase-16-trust-surface-hardening.md:613-634

Summary

One subcommand's contract in four layers — what it accepts, what it emits and what it claims to do. The manual documents a TGI output of tgi-launcher.sh + Dockerfile; the generator returns a single docker-compose document. Three auto-detection claims have no implementation, and the manual's "override any auto-detection with explicit YAML" contradicts its own Configuration section and cannot be implemented: ForgeConfig forbids extra keys and has no deployment: block.

Units

Unit Tracked by (October 2026 review)
F-W20260729-CLI-10 #259 · #260 · #435 · #483 · #612 · #618 · #786 · #815
F-W20260729-CLI-14 #259 · #260 · #483 · #567 · #612 · #618 · #786 · #815
F-W20260729-CLI-15 #259 · #260 · #483 · #612 · #618 · #705 · #786 · #815
F-W20260729-CLI-16 #80 · #259 · #260 · #483 · #612 · #618 · #786 · #815

Plan

  • Bind all four layers with one acceptance mechanism: golden tests comparing the manual's fenced blocks against real generate_deploy_config() output.

Exit criteria

  • Every deploy command in the manual runs in a temporary checkout, and the resulting tree and the quoted snippets match, in English and Turkish alike.
  • A negative test pins --output <existing dir> → success=False.
  • Absolute and relative local paths to HF Endpoints are config errors with no side effects.
  • A parametrized accepted/rejected flag matrix per target holds; TGI rejects port 65536; input + output budget ≤ the declared context window.
  • The generated compose file parses to a non-latest image and a read-only /data mount.

How to deliver

  • One root-cause family, committed on its own; then an Opus review round and a Sonnet review round, each with verified findings fixed and committed before the next step begins.
  • A reproducer that is red before the fix and green after; the full gauntlet from CLAUDE.md passes under ./.venv/bin/python.
  • Every English documentation edit ships with its Turkish mirror in the same commit; a new or renamed audit event gets its EN and TR catalog rows in the same PR.
  • A step that grows a deferred module carries its budget_history justification in the same diff; a widened guard rolls out non-strict first, then flips.
  • Commit bodies cite Absorbs F-W20260729-…; the CHANGELOG [Unreleased] section gets an entry only for a change a user or library consumer can observe (decision C-20).

Planned in docs/roadmap/phase-16-trust-surface-hardening.md (read at f94595f). Unit IDs (F-W20260729-…) come from the 2026-07-29/30 full-project review; the issue numbers next to them are the October 2026 review's records of the same defects.

主要言語
Python
スター
9
フォーク
1
平均マージ
7時間 42分
マージ済み PR（30日）
4

環境構築

はじめの一歩

  1. issue を最後まで読み、次にプロジェクトのコントリビューションガイドを読みます。
  2. 着手することを issue にコメントします — 二人が同じ作業をするのを防げます。
  3. リポジトリをフォークし、ブランチを切って変更します。
  4. issue 番号を参照したプルリクエストを送ります。

HodeTech/ForgeLM のほかの issue

HodeTech/ForgeLM の issue をすべて見る

似ている issue

Python の issue をもっと見る

新しい issue をメールで受け取る

初心者向けの GitHub issue を短くまとめたダイジェスト。