Teaching Claude Code to Write and Grow Its Own Skills: A Self-Replicating Agent Environment

A developer built a self-replicating agent environment for Claude Code that automatically discovers and accumulates reusable procedures as skills. Over several months, the system generated 64 auto-generated skills stored in ~/.claude/skills/auto/, using in-session generation and nightly batch harvesting with a curator to prevent overproduction and manage skill lifecycle.

Last time I wrote about splitting Claude Code's memory into four layers https://zenn.dev/bokuwalily/articles/f534402187bd07 . This is the follow-up: a setup where Claude Code itself discovers "reusable procedures" and accumulates them as skills over time. After running this for a few months, ~/.claude/skills/auto/ now holds 64 auto-generated skills. In this article I'll cover the design of that self-replication, the mechanisms that prevent overproduction, and the generation curve I observed while operating it. Claude Code is capable, but across sessions it forgets "how I got around this error last time." How to fix launchd 's exit 78, the steps to bake a favicon with png-to-ico , how to trim the RSC payload in Next.js — these procedures I should have already solved once were being rebuilt from scratch every single time. I could hand-write additions to CLAUDE.md, but that turns into a "whoever notices, writes it" workflow, and that doesn't last. So I flipped the idea around and decided to make Claude do the work of finding the procedures too . Skill generation is split into two tiers. | Tier | Timing | Cost | Owner | |---|---|---|---| | In-session generation | Any time during work | Free / instant | Claude itself, following CLAUDE.md instructions | | Nightly batch generation | Daily cron at 3:30 | Consumes Max quota | skill-harvest.sh claude -p headless | Then a Curator skill-curate.sh runs weekly to periodically organize the accumulated skills. The key is separating generation and organization into different processes. CLAUDE.md says this excerpt . スキル自己生成（auto-skills） 作業の中で再利用価値のある手順を見つけたら、頼まれなくても自分でスキル化する。 ただし乱造はしない。 生成トリガー（いずれか該当時のみ） - 5回以上ツールを使う非自明なタスクを最後までやり切った - エラー・行き止まりにぶつかった後、動く回避策を見つけた - ユーザーにアプローチを修正された（同じ修正を繰り返さないため） - 再利用できる手順・コマンド列・ワークフローを発見した The crux is " but don't overproduce ," plus narrowing the triggers down to four. Without writing this, Claude turns everything into a skill and the space fills up with noise almost immediately. Procedures that never got written down during a session get picked up from the conversation logs by a nightly batch. The skill-harvest.sh pipeline looks like this. .harvest-watermark the previous processing position raw/conversations/ .md claude -p <name /SKILL.md with three sections: frontmatter + Procedure / Pitfalls / Verification claude -p is the headless invocation I mentioned in the previous article. It runs on the Max quota, so API-key charges are zero. auto/ Auto-generated skills do not go in the same place as hand-written skills or plugin-bundled skills. They're pushed into an isolated namespace, ~/.claude/skills/auto/ . Two reasons. npx skills update -g only manages lock-registered skills. The ones in auto/ aren't added to the lock → auto/ and author: auto " → That author: auto is the Curator's target-detection key, and it's required in the frontmatter. --- name: <kebab-case description: <発火条件を具体的に author: auto ← Curatorの対象判定キー。必須 created: <YYYY-MM-DD version: 1.0.0 status: active active | stale（Curatorが自動設定） --- Left ungoverned, skills rot. The Curator runs weekly like this. .snapshots/ status: stale .archive/ .curator-proposals.md as :::message Not letting the Curator auto-delete is the core of the safe design. "Untouched for 30 days" doesn't necessarily mean the skill is bad you just haven't done that kind of work . So it stops at demotion and archival, leaving the final call to a human. ::: Looking at the distribution of created: , in the first three days after introduction 5/28–30 28 skills were verbalized in one burst, after which the pace dropped to a few per day. This matches intuition — the "unverbalized procedures" accumulated in the past get released early, and after that only newly encountered procedures get added. 12 created: 2026-05-30 9 created: 2026-05-29 7 created: 2026-06-09 6 created: 2026-06-11 5 created: 2026-06-10 5 created: 2026-05-28 Here's this morning's harvest log. 会話ログを精査した結果、以下の理由から該当なし。 - Codex CLI のセットアップ → codex-cli-setup が既存 - /insights を CLAUDE.md に反映 → insights-to-claude-md が既存 - カード債務の支払い優先順位 → 個人の財務状況に依存、再利用手順に非該当 該当なし 2026-06-15 04:11:21 harvest done exit 0, created=0 created=0 . It rejected duplicates of existing skills, rejected one-off content tied to personal circumstances, and terminated normally with 0 generated . This is proof that the "don't overproduce" design can process a miss as a correct miss. If it generated recklessly here, those 64 would balloon to 200 and turn to noise fast. From the Pitfalls sections of actual skills, the ones that mattered. claude -p hits the token cap when the full system prompt loads MAX THINKING TOKENS=10000 pmset wake on scheduleYou don't need the full pipeline. Just adding one block to CLAUDE.md gets Tier 1 running. ~/.claude/skills/auto/ author: auto in the frontmatter a marker for organizing later That alone means skills grow a little with every piece of work. You can add the nightly batch and the Curator later, once skills have accumulated and you feel like organizing them — that's soon enough. auto/ , and make author: auto the Curator's target keyNext time I'll write about how these skills grew too numerous and Claude Code itself got heavy — the audit that cut context injection from 228KB to 48KB https://zenn.dev/bokuwalily/articles/context-slimming . Lily @bokuwalily https://x.com/bokuwalily — solo developer. I build automation infrastructure with Claude Code while mass-producing iOS apps and web services. Your ❤️ and shares mean a lot Written by Lily — I ship iOS apps and automate my content stack with Claude Code. Follow along: Portfolio https://bokuwalily.com · X https://x.com/bokuwalily · GitHub https://github.com/bokuwalily