Kushandisa Agentic AI yeDevOps otomatiki neKiro uye LangChain
Introduction
Mumazuva ano anomhanya-mhanya kusimudzira software, zvikwata zveDevOps zvinogara zvichitsvaga nzira dzekuvandudza mashandiro uye kuvimbika kwekubatana kwavo kunoramba kuripo uye kuenderera mberi kwekuendesa (CI/CD) mapaipi. Imwe nzira inovimbisa ndeyekukwidziridza agent AI, iyo inogonesa kugadzirwa kwevakangwara vamiririri vanogona kuita otomatiki uye kukwidziridza mabasa eDevOps. Muchinyorwa chino, tichaongorora maitiro ekushandisa maAI-powered agents vachishandisa Kiro neLangChain, maviri ekucheka-kumucheto masimusi ari kushandura munda weDevOps otomatiki.
Chii chinonzi Agent AI?
Agentic AI inoreva rudzi rwehungwaru hwekugadzira iyo yakagadzirirwa kuita mabasa yakazvimiririra, senge mumiriri wemunhu. Muchirevo cheDevOps, agent AI inogona kushandiswa kuita otomatiki mabasa akadai sekuvaka, kuyedza, uye kuendesa software application. Nekukwirisa muchina kudzidza uye hunyanzvi hwekugadzirisa mutauro (NLP) maitiro, agentic AI vamiririri vanogona kudzidza kubva kune ruzivo uye kujairana nekuchinja nharaunda, vachiita kuti ive yakanaka kune yakaoma uye ine simba DevOps workflows.
Kusuma Kiro uye LangChain
Kiro uye LangChain maviri akavhurika-sosi masisitimu ayo akagadzirirwa chaizvo kuvaka agenti AI masisitimu. Kiro is a Rust-based framework inopa yakasimba uye scalable chikuva chekuvaka vakangwara vamiririri, nepo LangChain iri Python-based framework inotarisa paNLP nekunzwisisa mutauro.
**Kiro: A Rust-based Framework yekuvaka Intelligent Agents **
Kiro ndeyemazuva ano Rust-based framework inopa yakazara chikuva chekuvaka vakangwara vamiririri. NeKiro, vanogadzira vanogona kugadzira vamiririri vanogona kudyidzana neakasiyana masisitimu, masevhisi, uye data masosi, uye kuita mabasa akadai se:
- Automating workflows uye maitiro
- Kuongorora data uye kuita sarudzo
- Kukurukurirana nevanhu uye mamwe masisitimu
Kuvaka kwaKiro kwakagadzirirwa kuve kwakanyanya scalable uye kuchinjika, zvichiita kuti ive sarudzo yakanaka kune makuru-akawanda maDevOps otomatiki mapurojekiti.
**LangChain: Python-based Framework yeNLP uye Kunzwisisa Mutauro **
LangChain iPython-based framework inotarisa paNLP uye kunzwisisa mutauro. NeLangChain, vanogadzira vanogona kugadzira vamiririri vanogona kunzwisisa uye kugadzira mutauro wakafanana nevanhu, uye kuita mabasa akadai se:
- Kugadziriswa kwemutauro wechisikigo (NLP)
- Kuongororwa kwemavara uye ongororo yemanzwiro
- Shanduro yemutauro uye chizvarwa
LangChain's architecture yakagadzirirwa kuve yakanyanya modular uye inowedzera, zvichiita kuti ive sarudzo yakanaka yekuvaka tsika NLP uye kugona kunzwisisa mutauro.
Kuita AI-Powered Agents neKiro uye LangChain
Kuratidza simba reKiro neLangChain, ngatitarisei muenzaniso wakapfava wekuvaka AI-powered agent inogadzirisa pombi yeCI/CD. Mumuenzaniso uyu, isu tichashandisa Kiro kugadzira mumiriri anogona kudyidzana neGit repository, kuvaka uye kuyedza kodhi, uye kuendesa iyo application kupuratifomu yegore.
Heino kodhi kodhi yakapfava inoratidza maitiro ekugadzira Kiro mumiriri anogadzirisa pombi yeCI/CD:
import kiro
kubva ku langchain import LLAMA
mumiririri = kiro.Agent(
zita = "ci_cd_agent",
tsananguro = "Automate CI/CD pombi",
zviito=[
kiro.Action(
zita = "git_clone",
tsananguro = "Clones the Git repository",
code="""
import git
repo = git.Repo.clone("https://github.com/user/repo.git")
"""
),
kiro.Action(
zita = "kuvaka",
Description="Inovaka kodhi",
code="""
import subprocess
subprocess.run(["ita", "kuvaka"])
"""
),
kiro.Action(
zita = "test",
tsananguro="Inomhanyisa bvunzo",
code="""
import subprocess
subprocess.run(["ita", "test"])
"""
),
kiro.Action(
zita = "deploy",
tsananguro = "Inotumira iyo application",
code="""
import subprocess
subprocess.run(["ita", "deploy"])
"""
)
]
)
# Gadzira LangChain LLAMA modhi
llama = LLAMA(
zita = "llama_model",
tsananguro = "LLAMA modhi yekunzwisisa mutauro",
code="""
import langchain
muenzaniso = langchain.LLAMA(
zita = "llama_model",
tsananguro = "LLAMA modhi yekunzwisisa mutauro",
code="""
import langchain
muenzaniso = langchain.LLAMA(
zita = "llama_model",
tsananguro = "LLAMA modhi yekunzwisisa mutauro",
code="""
def process_text(zvinyorwa):
# Gadzirisa zvinyorwa uchishandisa iyo LLAMA modhi
return text
"""
)
"""
)
"""
)
# Gadzira Kiro mumiriri anoshandisa iyo LangChain LLAMA modhi
mumiririri = kiro.Agent(
zita = "ci_cd_agent",
tsananguro = "Automate CI/CD pombi",
zviito=[
kiro.Action(
zita = "git_clone",
tsananguro = "Clones the Git repository",
code="""
import git
repo = git.Repo.clone("https://github.com/user/repo.git")
"""
),
kiro.Action(
zita = "kuvaka",
Description="Inovaka kodhi",
code="""
import subprocess
subprocess.run(["ita", "kuvaka"])
"""
),
kiro.Action(
zita = "test",
tsananguro="Inomhanyisa bvunzo",
code="""
import subprocess
subprocess.run(["ita", "test"])
"""
),
kiro.Action(
zita = "deploy",
tsananguro = "Inotumira iyo application",
code="""
import subprocess
subprocess.run(["ita", "deploy"])
"""
)
],
language_model=llama
)
# Mhanya mumiririri
agent.run()
Iyi kodhi yekodhi inoratidza nzira yekugadzira Kiro agent iyo inogadzirisa pombi yeCI / CD uchishandisa iyo LangChain LLAMA modhi yekunzwisisa mutauro. Mumiririri anogona kudyidzana neGit repository, kuvaka uye kuyedza kodhi, uye kuendesa iyo application kupuratifomu yegore.
Pamusoro 5-10 Inonyanya Kushandiswa Mirairo/Zvidimbu
Heano epamusoro 5-10 anonyanya kushandiswa mirairo/zvimedu zveKiro neLangChain:
kiroutil init: Inotangisa chirongwa chitsva cheKiro.kiroutil kuvaka: Inovaka Kiro mumiririri.langchain init: Inotanga chirongwa chitsva cheLangChain.langchain kuvaka: Inovaka muenzaniso weLangChain.kiroutil run: Inoshandisa Kiro mumiririri.langchain bvunzo: Miedzo muenzaniso LangChain.kiroutil deploy: Inotumira Kiro mumiririri kune gore papuratifomu.langchain evaluate: Inoongorora muenzaniso weLangChain.kiroutil debug: Debugs a Kiro agent.langchain kuona: Visualize a LangChain modhi.
Kuona
Heino ratidziro yeKiro neLangChain architecture uchishandisa Chart.js:
{
"type": "bha",
"data": {
"labels": ["Kiro", "LangChain"],
"datasets": [
{
"label": "Architecture",
"data": [100, 80],
"backgroundColor": ["#4CAF50", "#FF9800"]
}
]
},
"sarudzo": {
"zita": {
"kuratidza": chokwadi,
"zvinyorwa": "Kiro uye LangChain Architecture"
},
"legend": {
"kuratidza": chokwadi
}
}
}
Ichi chitarisiko chinoratidza zvivakwa zvepamusoro zveKiro neLangChain, zvichiratidza masimba avo uye kushaya simba.
Mhedziso
Muchikamu chino, takaongorora simba re agent AI yeDevOps automation uchishandisa Kiro uye LangChain. Nekusimudzira aya ekucheka-kumucheto masisitimu, vagadziri vanogona kugadzira vakangwara vamiririri vanogona kuita otomatiki uye kukwenenzvera CI/CD mapaipi, kuvandudza mashandiro esoftware uye kuderedza njodzi yekukanganisa. NeKiro neLangChain, vagadziri vanogona kugadzira tsika NLP uye kugona kunzwisisa mutauro, uye kudyidzana neakasiyana masisitimu, masevhisi, uye data masosi. Isu takapawo kodhi yekodhi yakapfava inoratidza nzira yekugadzira Kiro agent iyo inogadzirisa pombi yeCI / CD uchishandisa LangChain LLAMA modhi yekunzwisisa mutauro.