agentsclimarketplace

Skill creator safe

Skill alidevai/chiron_agent/.claude/skills/skill-creator-safe

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 skill-creator-safe

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

Aranan skill guvenilir kaynaklarda bulunamazsa, birincil kaynaklardan (resmi dokumantasyon, standartlar, akademik makaleler) yeni bir skill uretme prosedurudur. Uretilen skill her zaman experimental baslar ve normal edinme akisindaki tarama/eval/onay kapilarindan gecer.

SKILL.md

2.5 KB, as published. Nobody here has run it

Skill Creator (Safe)

Hazir skill bulunamadiginda son caredir. Agent yalnizca kendi genel bilgisine dayanarak YUKSEK RISKLI skill uretmez; birincil kaynaklara dayanir.

Kaynak onceligi (zorunlu)

  1. Resmi dokumantasyon
  2. Standartlar (RFC, spec, resmi rehber)
  3. Akademik makaleler
  4. Guvenilir acik kaynak projelerin kodu
  5. Alan uzmanlarinin iyi bilinen rehberleri
  6. Projenin kendi kurallari

Her prosedur adimi en az bir kaynaga baglanmalidir. Kaynaksiz "genel bilgi" yuksek riskli alanlarda (finans, guvenlik, saglik) tek basina yeterli degildir.

Uretim akisi

Alan arastirmasi -> birincil kaynak toplama -> uzman is akisi cikarma
  -> SKILL.md taslagi -> ornekler + karsi ornekler -> test fixture'lari
  -> evals/<id>.yaml -> staging'e alma -> tarama -> eval -> experimental kayit

Uretilen SKILL.md sablonu

Her uretilen skill su bolumleri icermelidir:

---
name: <ad>
version: 0.1.0
status: experimental
description: <ne zaman, hangi amacla>
domains: [<alan>]
risk_level: <low|medium|high>
required_tools: [...]
allowed_permissions: [...]
forbidden_permissions: [...]
sources:
  - url: <birincil kaynak>
    type: official_documentation
evaluation_suite: evals/<ad>.yaml
---

# Amac
# Kullanim kosullari
# Kullanilmamasi gereken durumlar
# On kosullar
# Zorunlu prosedur
# Kalite kapilari
# Guvenlik kurallari
# Hata durumlari
# Cikti formati
# Ornekler
# Testler
# Kaynaklar

Uretim sonrasi (atlanmaz)

  1. evals/<id>.yaml olustur: en az bir olculebilir kontrol (frontmatter, guvenlik taramasi, mumkunse calisan bir ornek).
  2. python -m core stage <taslak-dizin> --id <ad> --risk <seviye> --source generated
  3. security-gatekeeper ve evaluator subagent'larindan bagimsiz inceleme al.
  4. python -m core eval <id> -> gecerse sandbox-validated.
  5. Uretimi yapan agent nihai onaylayici DEGILDIR; promote/approve akisina birak.

Olgunluk seviyeleri

experimental -> sandbox-validated -> project-approved -> production-approved

Her yeni uretilen skill experimental baslar; kullanim ve dogrulama ile terfi eder.

Kaynaklar

  • Ilgili: [[secure-capability-acquisition]], [[capability-gap-analysis]]

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.