Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/aiya000/dotfiles/save-memorynpx skills add aiya000/dotfiles --skill save-memorygit clone --depth 1 https://github.com/aiya000/dotfilesWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00034 | $0.01158 |
| Opus 5 | $0.00017 | $0.00579 |
| Sonnet 5 | $0.00007 | $0.00232 |
| Haiku 4.5 | $0.00003 | $0.00116 |
Grade A, and why
save-memory scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 91 lines — stays where its author put it; the contents beside it link to each section on GitHub.
save-memory
What you do with this skill
This skill distills the current conversation into a compact, machine-readable memory file that future Claude Code sessions can load to restore context.
Steps
-
Use the current conversation content as source material
-
Determine current date and time by running
date +%Y-%m-%d_%H:%M -
Verify the memory directory is set up, using the Bash tool:
- Run:
ls -ld ~/.ai-memoryand confirm the output starts withl(i.e.~/.ai-memoryis a symlink) - If it is not a symlink (or does not exist), stop and ask the user to set it up. Do not create it yourself
- Otherwise, list existing files with:
ls ~/.ai-memory 2>/dev/null
- Run:
-
Secret scan — before writing, review the content you are about to save for potential secrets or sensitive values:
What to look for:
- API keys and tokens: strings starting with
ghp_,gho_,AKIA,sk-,xox, or matching-----BEGIN.*PRIVATE KEY - Variables with sensitive names holding a value: patterns like
API_KEY=,_SECRET=,_TOKEN=,PASSWORD= - Hardcoded absolute home paths:
/Users/<name>/or/home/<name>/(prefer~) - Personal or organizational proper nouns that could identify specific individuals or organizations
If any of the above are found in the content to be saved, use AskUserQuestion to present these options:
- "Save as-is" — proceed despite the finding
- "Fix then save" — let the user review and fix the content first
- "Cancel" — abort without writing
Only proceed with writing if the user selects "Save as-is", or if nothing suspicious was found.
- API keys and tokens: strings starting with
-
Write (or append) to:
~/.ai-memory/YYYY-MM-DD-{project}-{topic}.md- If a file for the same or similar project+topic already exists: append or update it
- Otherwise: create a new file with header
# Memory - YYYY-MM-DD {project}is the short name of the current project (e.g.dotfiles,my-app); omit if the topic is not project-specific
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 91 lines · 34 tokens per session scan A 9b27fe64b4c7
save-memory is a skill published in the GitHub repository aiya000/dotfiles (19 stars, last pushed 2d ago), licensed MIT. It adds 34 tokens to every session and 1,158 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
coding
Principles for writing and designing code, covering API and abstraction design, naming, comment discipline, and standards of evidence for claims about code behavior. Use whenever the task is to write or modify code, design an API or software architecture, or review code.
restart__pull_request
乱れた (fix-up コミット過多・議論の脱線・大規模な書き直し) プルリクエストを、 ユーザーが "restart" / "やり直し" したいときに起動する。既存 PR を close し、 ブランチを新ブランチ上の単一のクリーンなコミットに squash し、過去の PR 議論 (レビューコメント・スレッド・リアクション) を統合した説明文で新しい PR を作成する。 元のブランチを force-push してはならない。必ず新しいブランチを作成する。.
validate__japanese
日本語の Markdown 文書 (README・ドキュメント・ブログ下書き・.md) を編集した後に起動する。textlint の ja-technical-writing / ja-spacing プリセットで、長すぎる一文・読点過多・文体の混在・全角半角スペースを検出する。さらに文中ハードラップ (文末でない位置で折り返した、意味のない改行) を検出し、段落・箇条書き項目を 1 行に連結して調整する。1 行に文を積み上げた箇所は、箇条書きへの変換候補として検出する。実在のシンボルを指すコードスニペットへ参照リンクも付与する。日本語の .md を変更したときに使用する。.
prepare__issue
あなたは要件定義済みの Issue を実装着手可能な状態へ引き上げるオーケストレーターである。 対象は status:acknowledged の Issue であり、完了時に status:ready へ遷移させる。.
rescue__pull_request_review
GitHub PR にレビュアー (人間または AI) がコメントを残した後に起動する。 レビューコメントを読み取り、要求されたコード変更を適用し、コミットして push する。 レビューフィードバックの各ラウンドごとに繰り返す。ユーザーの確認は不要 — Claude がこのプロセス全体を自律的に起動・実行する。.
submit__pull_request
プルリクエストを作成から完了まで一貫して提出するときに起動する。ナラティブ型の PR 説明文を生成し、PR を作成し、CI チェックを監視し、CI の失敗を自動修正する。 PR ナラティブと CI 修正のワークフローを 1 つの自律フローに統合する。.