lilmgenius/paperthin

detool

Make durable artifacts portable by replacing incidental stack, vendor, model, CLI, path, quota, or harness nouns with the mechanism they mean, while keeping provenance, runbooks, and tool-subject claims concrete.

ソースを見る

使い方

  1. インストールコマンドをコピーし、自分の環境へ導入する前にソースを確認します。
  2. 必要なツール、認証情報、対応環境を原文で確認します。
  3. 入力と成果物を指定し、出力を資料と照合します。
npx skills add https://github.com/lilmgenius/paperthin --skill detool

開始前の確認

  • 導入にはモデル契約、API枠、ファイルへのアクセス許可は含まれません。
  • スター数や導入数は活動の指標であり、独立した品質・安全性試験ではありません。
リポジトリの原文

見出し、例、コード、表、リンク、参照画像を含む原文を表示しています。

Remove incidental tool coupling from artifacts that claim to outlive the stack that wrote them.

Goal

A durable artifact should stay true and actionable when the agent, harness, operating system, vendor, model, or toolchain changes. detool finds stack-specific nouns in portable content and rewrites them to the underlying mechanism without deleting the action they enabled.

The point is portability, not vagueness. A reader should still know what to inspect, run, measure, compare, or verify after the rewrite.

Workflow

  1. Read the target artifact end to end.
  2. Role before edit: classify the whole artifact, or each section if mixed, as durable/portable content, provenance/operational record, or a claim about a named tool.
  3. Sweep durable content for incidental coupling: harness paths, vendor CLIs and flags, model or product brands used as mechanisms, tool-specific environment variables, quotas, UI steps, cache homes, session files, and version-pinned behavior stated as timeless truth.
  4. Mechanism not euphemism: replace each durable-content hit with the neutral mechanism that preserves the action. If neutral wording would lose the action, keep the concrete detail as an example of the mechanism instead of pretending it generalizes.
  5. Keep provenance and operational content concrete. Build records, capsules, benchmark logs, install guides, runbooks, tool-targeted how-tos, command transcripts, and exact reproduction steps are supposed to name the stack that produced them.
  6. Keep comparative and tool-subject claims concrete. If the sentence is about a named tool, vendor, model, bug, prior-art source, or measured limit, the name is the subject, not incidental coupling.
  7. Re-read as a reader on a different stack: the artifact should still be true, portable, and executable where it promised action.
  8. Report what was neutralized, what was deliberately kept, and the role judgment behind each keep.

Rules

  • Role before edit. Never scrub before deciding whether the text is portable content, provenance, operational instruction, or a tool-subject claim.
  • Mechanism, not euphemism. "Use the session execution log and sum usage records" is a mechanism; "check the logs" is a euphemism that loses the work.
  • Preserve actionable detail. Abstract only the coupling, not the procedure, evidence, comparison, or constraint.
  • Keep provenance and operational records exact. Concrete paths, commands, flags, versions, products, and UI steps stay when they record what happened or explain how to drive a specific tool.
  • Keep comparative/tool-subject names. A bug report about a CLI, a benchmark against a model, a prior-art citation, or a vendor-specific limit keeps the named subject.
  • A pass that finds no incidental coupling changes nothing.
  • Mutate with edit-safety: assert each target exists and report a MISS instead of silently no-oping; edit unicode-safe; replace per occurrence, never by blanket sweep; script large structural moves.

Verification

Before finishing, confirm:

  1. Every changed hit was in durable or portable content.
  2. Every rewrite preserved the actionable mechanism.
  3. Every kept tool noun is provenance, operational instruction, or the subject of its claim.
  4. The report names neutralized couplings, deliberate keeps, and any judgment calls.

タスクに関連する項目

同じリポジトリから

関連する Skills

すべての Skills
lilmgenius
コミュニティ

aim

When data lands with a thin ask — files, a repo, a paste, a chat scrollback — read it and propose what the user is trying to do with it, as a proposal to confirm rather than a question to answer. Use the moment material arrives without a fully-formed ask, or when the ask reads thinner than the data suggests.

1時間の導入数
1
GitHub Stars
1300
更新日
10月20日
lilmgenius
コミュニティ

feynman

Press a decision you just made until you can explain it to a skeptic in your own words, or surface the gap where you can't. The Feynman test aimed at a choice: understanding is cheapest to fake the moment after you decide, so a fresh critic presses for what you can't actually explain. Use right after picking an option, especially one an agent laid out without arguing for.

1時間の導入数
1
GitHub Stars
1300
更新日
10月20日
lilmgenius
コミュニティ

hate

Attack a plan, design, or argument like you want it to fail before you commit real effort — return the single load-bearing objection and the cheapest experiment that would prove it matters, not a checklist. User-invoked on purpose: a hate-it reflex always in the agent's reach would bias it toward demolition. Weighs many failure axes but owns the synthesis to one root.

1時間の導入数
1
GitHub Stars
1300
更新日
10月20日
lilmgenius
コミュニティ

macrothink

User-invoked read-only pass for checking whether the current direction is tunnel-visioned: strip the session's bait, fan out 2 to 5 same-model fresh reads, and report divergence first without treating convergence as proof.

1時間の導入数
1
GitHub Stars
1300
更新日
10月20日