Skip to content
Agentmart
Submit a tool

process-learning

Learning Eventが実際に発生したtaskだけ、成果物の問題とループ中の無駄・誤判断を振り返り、telemetryを再利用可能な改善Evidenceとして使う。Risk R3/R4だけを理由に起動しない。

Skill by hondasports

IndexedMIT
Write a review

Chalk Labs

Need help deploying process-learning?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/hondasports-process-learning

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Learning Eventが実際に発生したtaskだけ、成果物の問題とループ中の無駄・誤判断を振り返り、telemetryを再利用可能な改善Evidenceとして使う。Risk R3/R4だけを理由に起動しない。

Use cases: productivity

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/process-learning

What's inside

Skill name
process-learning
Description
Learning Eventが実際に発生したtaskだけ、成果物の問題とループ中の無駄・誤判断を振り返り、telemetryを再利用可能な改善Evidenceとして使う。Risk R3/R4だけを理由に起動しない。
License
MIT
Instructions
148 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Process Learning

## Event-driven only

次のEventが1つ以上ある時だけ起動する。

- human correction
- unexpected Gate / CI / E2E failure
- actionable review finding
- repeated retry / Incident
- scope / impact miss
- delivery / aftercare miss
- process rule / enforcement不足が明確になった

Eventなし:

```text
Learning event: none
Status: NOT_REQUIRED
```

Risk R3/R4だけでは起動しない。

## Telemetryを先に見る

`task-state.telemetry` がある場合、full logやchatを読み返す前にstage summaryを見る。

- PREPARE / IMPLEMENT / VERIFY / REVIEW / DELIVER / AFTERCARE elapsed
- external wait
- source reads / skill loads
- retries / full suite runs / review cycles
- changed files / AC / IV / TC / Controls / findings数
- Spec Confidence / Risk / max observed Risk / task size

時間単体で良し悪しを決めない。同程度の規模・Risk・Spec Confidenceに対して、どこへ時間・retry・再読込が偏ったかを見る。

例:

- PREPARE長い + source_reads多い → discoveryやcompact contractを疑う
- VERIFY長い + full_suite_runs複数 → fail-fast順 / CI evidence reuseを確認
- REVIEW長い + review_cycles多い → omission scanやcontract不足を確認
- AFTERCARE長い + external_wait比率高い → Agent loop自体の遅さと混同しない
- IMPLEMENT長い + requirements_gap → PREPAREで早期検出できたか確認

wall-clockにはtool/API/CI待ちが含まれる。分離できない時間を推測しない。

Token usageはruntimeが正確に提供する時だけ補助Evidenceにする。

## 分析

```text
Observed problem:
Immediate cause:
Process cause:
Why existing enforcement did not catch it:
Earlier detection / prevention:
Reusable rule:
```

## Loop retrospective

目的は3つだけ。

- `context`: 読み込む情報・保持状態・重複説明を減らす
- `speed`: tool round-trip、待機、重複実行、手戻りを減らす
- `precision`: scope miss、requirements gap、test gap、false completionを減らす

新たに全ログを読み直さず、task-state / telemetry / findings / Verification Evidence / 既存tool結果を使う。

影響の大きいものを最大3件まで。

- 無駄な判断・行動
- 間違った判断
- 維持すべき判断

hindsightだけで言えるものと、当時のEvidenceから回避可能やったものを分ける。改善候補は回避可能なものを優先する。

改善は「手順追加」より先に:

- 削除
- 統合
- delayed loading
- 順序変更
- Evidence reuse
- cheap deterministic check

を検討する。Required Controlや品質Gateを速度目的で弱めない。

## Target priority

1. Script / code
2. CI / deterministic enforcement
3. Skill
4. AGENTS policy
5. Runbook / docs

既存ruleを破っただけなら文章追加よりenforcement改善を優先する。

## Result persistence

再利用可能な候補をchat上の提案だけで終わらせない。

各候補に:

- observed problem / process cause
- reusable rule
- improvement axes
- proposed target
- disposition
- evidence

を記録する。

Disposition:

- `applied`: loop artifactへ反映済み。location + verification evidence必須
- `follow_up`: current task scope外。Issue / task / PRのtype・reference + target + rationale必須
- `no_change`: 既存enforcementで充足済み、または再利用不能。rationale + evidence必須

From hondasports/Re-Me, MIT.

More from hondasports/Re-Me

Signals

Stars
0
Forks
0
Open issues
2
Last commit (this folder)
Aug 28, 2026 (2 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from hondasports

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get process-learning running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying process-learning?

CallGet help
Cite this page

process-learning skill. Agent Mart. https://agentmart.dev/skills/hondasports-process-learning (last updated 2026-10-24)