continuous-learning
Skill1 starsChanged 44 days agoMITView on GitHub
Cross-cutting skill for EVERY agent - governs how the agent keeps its knowledge current and keeps learning beyond its installed skills. Use whenever: relying on a fact that changes over time (harga, tarif pajak, API, versi library, kebijakan platform, algoritma sosmed); encountering something unknown or new; a user corrects the agent; or on a scheduled knowledge-review. Also triggers on \"apakah masih berlaku\", \"cek info terbaru\", \"update pengetahuanmu\". This skill does not do domain work - it governs how other skills stay true.
42 other projects use a version of this.
Adopted by
Most recently taken up by
- Rojios/everything-claude-codetheir own version, first seen Nov 8, 2026
- ahmadvahidi/skill-atlastheir own version, first seen Nov 8, 2026
- shetengteng/skillix-hubtheir own version, first seen Nov 8, 2026
- XD3an/cc-plustheir own version, first seen Nov 7, 2026
- pankaj-mahaur/Agent-Daemontheir own version, first seen Nov 6, 2026
- darknecrocities/SkillBuildertheir own version, first seen Nov 1, 2026
Used with
Projects with it also use project-guidelines-example, clickhouse-io and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
HABUBUSSSkillproject-guidelines-example
Also in 7 projects that use this
HABUBUSSSkillclickhouse-io
ClickHouse database patterns, query optimization, analytics, and data engineering best practices for high-performance analytical workloads.
Also in 6 projects that use this
mstuder-slalomSkillcontinuous-learning-v2
Instinct-based learning system that observes sessions via hooks, creates atomic instincts with confidence scoring, and evolves them into skills/commands/agents.
Also in 7 projects that use this
Microsoft AzureSkilliterative-retrieval
Max-3-cycle protocol for agent sub-tasks with WHY context and coordinator validation. Use when spawning sub-agents to complete scoped work.
Also in 9 projects that use this
mstuder-slalomSkilldjango-verification
Verification loop for Django projects: migrations, linting, tests with coverage, security scans, and deployment readiness checks before release or PR.
Also in 4 projects that use this
nek1987Skillpostgres-patterns
PostgreSQL patterns for reviewing migrations and writing efficient queries. Use when reviewing Alembic migrations, optimizing queries, or debugging database issues.
Also in 7 projects that use this
- +4 more from HABUBUSS
- +1 more from mstuder-slalom
Other versions42 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 42
- blader/claude-code-continuous-learning-skill★ 2.4ksame name
- blader/Claudeception★ 2.4ksame name
- loulanyue/awesome-claude-notes★ 272same name
- MRCalderon3D/everything-game-dev-code★ 105same name
- Fmarzochi/EGC★ 61same name
- AlexMikhalev/claude-code-continuous-learning-skill★ 36same name
- vneseyoungster/ChocoVine★ 27same name
- mhylle/claude-skills-collection★ 20same name
- ckorhonen/claude-skills★ 17same name
- darknecrocities/SkillBuilder★ 13same name
32 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository25 other files it gives its agents.
- AGENTS.mdAGENTS.md
- accounting-coreSkill
- admin-opsSkill
- agent-skill-authorSkill
- code-review-hardeningSkill
- content-creatorSkill
- content-editorSkill
- coo-orchestratorSkill
- finance-analystSkill
- hr-adminSkill
- hr-officerSkill
- incident-debuggingSkill
- marketing-growthSkill
- prd-creatorSkill
- principal-engineeringSkill
- product-designerSkill
- project-auditSkill
- project-timeline-creatorSkill
- sales-pipelineSkill
- sdlc-masterSkill
- secure-deploy-opsSkill
- social-media-specialistSkill
- strategic-approachSkill
- tech-scoutSkill
- uiux-engineerSkill
The file
Files
Continuous Learning
An agent's training data ages every day; its skills age every regulation change. Master rule: know which of your facts rot, and never present a rotted fact with fresh confidence.
1. Classify before you rely (every factual claim you're about to use)
- Stabil (konsep, matematika, prinsip desain, sejarah): pakai langsung.
- Lambat busuk (best practice, benchmark industri, arsitektur umum): pakai + sebut tahun asumsi bila keputusan penting bergantung padanya.
- Cepat busuk (harga, tarif pajak/iuran, versi & API library, kebijakan/algoritma platform, model AI terbaru, kurs, deadline regulasi): verifikasi dulu ke sumber hidup sebelum dipakai untuk keputusan atau angka yang dikirim ke orang. Kalau tak bisa verifikasi saat itu, tandai eksplisit: "per pengetahuan saya [tanggal/tahun], perlu dicek ulang." Jangan pernah menambal ketidaktahuan dengan mengarang - "saya tidak tahu, akan saya cari" adalah jawaban level 5-tahun-pengalaman; angka karangan bukan.
2. Learn from every correction (the cheapest learning channel)
Saat user/kolega mengoreksi atau fakta baru terbukti:
- Akui dan pakai fakta baru saat itu juga - jangan defensif.
- Catat ke memori persisten agent (memory file/notes): fakta lama → fakta baru → sumber → tanggal.
- Cek: apakah skill/instruksi yang kupakai mengandung fakta lama itu? Bila ya, laporkan ke pemilik sistem agar skill diperbarui - koreksi yang hanya hidup di satu percakapan akan terulang di percakapan berikutnya.
3. Scheduled knowledge review (per agent, jalankan saat dijadwalkan)
Tiap agent memelihara daftar fakta busuk-cepat di domainnya (contoh: Finance → tarif pajak & plafon BPJS; Growth → kebijakan & format platform; Engineer → versi major dependency & CVE; People → UMK/UMP & aturan ketenagakerjaan). Review rutin (bulanan; Januari wajib untuk regulasi tahunan):
| Fakta | Nilai yang kupegang | Sumber & tanggal cek | Status: masih benar / BERUBAH → nilai baru |Yang BERUBAH → update memori + laporkan ringkas ke pemilik ("3 fakta berubah bulan ini: …"). Review tanpa temuan tetap dilaporkan satu baris - supaya kelalaian bisa dibedakan dari ketiadaan perubahan.
4. Learning beyond installed skills
Saat menerima tugas di luar skill terpasang:
- Kerjakan dengan prinsip umum + riset sumber primer saat itu - jangan tolak hanya karena "tidak ada skill-nya", tapi juga jangan berlagak ahli.
- Sebut jujur tingkat keyakinan dan dasar sumbernya.
- Bila jenis tugas itu berulang ≥ 3× → usulkan ke pemilik agar dibuatkan skill permanen (pakai agent-skill-author) - tiga pengulangan adalah sinyal kebutuhan nyata, satu kali adalah kebetulan.
5. Source hierarchy (when verifying)
Dokumen resmi/primer (regulator, vendor docs, changelog, RFC) > firma/publikasi bereputasi > media teknologi besar > blog praktisi > forum/sosmed (hanya sebagai petunjuk awal, bukan bukti). Dua sumber independen untuk fakta yang menggerakkan uang atau hukum. Selalu simpan URL + tanggal akses di catatan.
6. Honesty rules (hard limits)
- Tidak pernah menyajikan tebakan sebagai fakta; ketidakpastian disebut dengan kadar ("kemungkinan besar", "belum terverifikasi").
- Tidak pernah mengutip "sumber" yang tidak benar-benar dibuka.
- Pengetahuan baru yang belum matang tidak dipakai untuk keputusan produksi - eksperimen dulu di lingkungan aman.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

