agentsclimarketplace

Capability gap analysis

Skill alidevai/chiron_agent/.claude/skills/capability-gap-analysis

Autonomous, self-improving AI agent platform for Claude Code — safe capability & skill acquisition, MCP/tool discovery, sandboxed policy gates, and token-efficient learning.

Install
npx -y skills add alidevai/chiron_agent --skill capability-gap-analysis

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

3 things to look at

  • 20 days oldThe repository was created 20 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Bir goreve baslamadan once yetkinlik acigini ve guven skorunu olcer; eksik olanin bilgi mi, skill mi, tool mu oldugunu ayirir ve sonraki adimi (dogrudan yurut / dogrulayarak yurut / yetenek edin) belirler. Yuksek riskli, belirsiz veya "en guncel/en iyi/dogrula" istenen gorevlerde ilk adim olarak kullanilir.

SKILL.md

2.6 KB, as published. Nobody here has run it

Capability Gap Analysis

Bu skill, "her seyi bildigini varsayan" davranisin panzehiridir. Goreve dalmadan once sistemin o gorev icin gercekten hazir olup olmadigini kanitla olcer.

Ne zaman kullanilir

Su durumlardan HERHANGI biri varsa, koda/analize baslamadan once bunu calistir:

  • Alan yuksek riskli: finans, trading, guvenlik, saglik, hukuk
  • Kullanici "en guncel", "en iyi", "dogrula", "emin ol" diyor
  • Gorev mevcut skill'lerin acikca kapsami disinda
  • Kullanilacak kutuphane/standart guncel olmayabilir
  • Ayni hata tekrar ediyor veya bir tool cagrisi iki kez basarisiz oldu
  • Kullanici ozel/bilinmeyen bir teknoloji, format veya sistem istiyor

Prosedur

  1. Gorevi alanlara ayir. task_domain ve task_subdomains cikar. Ornek: "Solidity kontrati guvenlik incelemesi" -> domain=security, subdomains=[solidity, smart-contract, static-analysis]

  2. Gerekli yetenekleri listele. Hangi prosedurler (skill), hangi araclar (tool/MCP), hangi veri kaynaklari gerekiyor? Somut skill id'leri olarak yaz.

  3. Gap raporunu calistir:

    python -m core gap --domain <alan> --skills <id1,id2,id3> --risk <low|medium|high|critical>

    Cikti: existing / missing / stale yetenekler, confidence (0-1), action.

  4. Action'a gore yonlen:

    • proceed (confidence >= 0.75, eksik yok): dogrudan mevcut skill'lerle yurut.
    • proceed_with_verification (0.60-0.75): yurut, ama sonucu evaluator subagent'ina bagimsiz dogrulat. Yuksek riskte bu zorunludur.
    • research_and_stage_capability (< 0.60 veya eksik var): capability-manager subagent'ini cagirarak eksik yetenegi guvenli sekilde edin (bkz. [[secure-capability-acquisition]]).

  5. Riske gore kapilar:

    • risk high/critical ise: bagimsiz evaluator ZORUNLU.
    • Eylem geri donusu zor ise (production, email, trade, dosya silme): insan onayi.

Kullanilmamasi gereken durumlar

  • Onemsiz, tek seferlik, dusuk riskli gorevlerde (or. bir yazim hatasi duzeltme) bu analiz gereksiz surtunmedir; dogrudan yap.

Cikti formati

Gorevin basinda kisa bir "hazirlik notu" ver: alan, gerekli yetenekler, confidence skoru, secilen action ve gerekiyorsa hangi subagent'in cagrilacagi.

Kaynaklar

  • Platform gap motoru: core/confidence.py
  • Ilgili: [[secure-capability-acquisition]], [[skill-router]]

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.