Ukufika I-Xcode 26.3 kuphawula uguquko ekuthuthukisweni kwesofthiwe ku imvelo ye-apulaLe nkampani inqume ukuhlanganisa ngqo ama-ejenti obuhlakani bokwenziwa avela ku-Anthropic kanye ne-OpenAI ngaphakathi kwe-IDE yayo esemthethweni, ithatha isinyathelo esicacile sokuya kulokho okubizwa ngokuthi i-ejenti yokufaka amakhodi ngaphandle kokuphoqa amaqembu ukuthi ayeke imisebenzi yawo evamile yokusebenza.
Lo mkhankaso uthakazelisa kakhulu ama-studio kanye nama-startups e- ISpain kanye neYurophu yonkelapho intuthuko I-iOS, i-iPadOS, ne-macOS Kuyindawo enokuncintisana okukhulu kodwa enomkhawuko wezinsizakusebenza. Nge-Xcode 26.3, i-Apple iphakamisa ukuthi imisebenzi eminingi yokuhlela ephindaphindwayo neyinkimbinkimbi idluliselwe kuma-ejenti e-AI, kuyilapho unjiniyela egcina ukulawula kokugcina kuzo zonke izinguquko.
Kusho ukuthini ngempela ukufaka amakhodi e-ejenti ku-Xcode 26.3?
La ikhodi ye-ejenti Kudlula kakhulu ukuqedela okuzenzakalelayo kwakudala. Ku-Xcode 26.3, ama-ejenti e-AI avela ku- I-Anthropic (Claude Agent) kanye I-OpenAI (I-Codex kanye nezinhlobo ze-ChatGPT ezisekelwe kukhodi) zingathola imiyalelo ngolimi lwemvelo, ukuhlaziya iphrojekthi, ukuthuthukisa uhlelo lokusebenza, kanye nokwenza uchungechunge lwezinyathelo ngaphandle kokuthi unjiniyela aqondise yonke into eyenzeka.
Empeleni, lokhu kusho ukuthi umhleli wezinhlelo angacela into ethile efana nokuthi “Engeza isikrini sokungena ngemvume ngokuqinisekiswa kwe-biometric"futhi uvumele i-ejenti ukuthi inakekele ukuthola amamojula afanele, ukudala noma ukuguqula amafayela, ukulungisa izilungiselelo zephrojekthi, ukuhlanganisa, ukusebenzisa izivivinyo, kanye nokubuyekeza umphumela ngaphambi kokubuyisela ukulawula."
Isihluthulelo ukuthi i-Xcode idalula amakhono ayo angaphakathi—umhlanganisi, uhlelo lokuhlola, ukubuka kuqala, imibhalo—njengesethi ehlelekile yamathuluzi i-ejenti engawafinyelela. Ngakho-ke, i-AI ayikhiqizi nje kuphela ikhodi ehlanganisayo, kodwa futhi iyakwazi qinisekisa ukuziphatha kanye nesixhumi esibonakalayo ngokusebenzisa ukubuka kuqala kwe-SwiftUI kanye nezinye izindlela ze-IDE uqobo.
Ngokungafani nezinye izindlela ezingacaci kahle, i-Apple ikhetha ukusebenzisa lapho unjiniyela engabona khona ukuthi i-ejenti yenzani ngaso sonke isikhathi: Isenzo ngasinye siyaqoshwa futhi yonke ibhulokhi yezinguquko iyagqanyiswa. kumhleli, kunciphisa umuzwa "webhokisi elimnyama" odala ukungabaza okuningi emaqenjini asebenza ngekhodi ebalulekile.
Yini efakwa yi-Anthropic kanye ne-OpenAI agents ku-IDE?
Nge-Xcode 26.3, onjiniyela bangafaka ama-ejenti ngokuchofoza okukodwa. I-Agent kaClaude (i-Anthropic) y I-Codex/ChatGPT (OpenAI) kusuka kuphaneli yezilungiselelo ze-IDE uqobo. Uma usulandiwe, vele ungene ngemvume noma ufake Ukhiye we-API ukuze bakwazi futhi baqale ukusebenza nabo kusuka ebhokisini lombhalo eliseceleni, njengokungathi ingxoxo evumelaniswe nephrojekthi.
Lawa ma-ejenti ayakwazi Hlola isakhiwo sephrojekthi kanye nemethadatha yayoukuthola amafayela afanele ngokushesha, ukuhumusha indlela ikhodi ehlelwe ngayo, nokuhlonza ukuncika okuhilelekile. Ukusuka lapho, bangakwazi:
- Dala imisebenzi emisha, amakilasi, noma amamojula aphelele ngokulandela incazelo ngolimi lwemvelo.
- Shintsha ikhodi ekhona ukuze ivumelane nayo Ama-API e-ecosystem akamuva e-Apple noma izinhlaka ezintsha.
- Lungiselela futhi usebenzise amabhethri ukuhlolwa kweyunithi nokuhlanganiswa ukuqinisekisa izinguquko.
- Phindaphinda amaphutha atholakele, usebenzise izilungiso futhi uhlanganise kabusha ngaphandle kokungenelela okuqhubekayo.
- Khiqiza izincazelo zesinyathelo ngesinyathelo zalokho abakwenzile nokuthi kungani, okuwusizo ekubukezeni nasekuqeqesheni iqembu.
Isici esisodwa esibaluleke kakhulu ukufinyelela ngqo kwama-ejenti ku- Imibhalo esemthethweni neyakamuva ye-AppleEsikhundleni sokuthembela kuphela kulokho ekufundile ngesikhathi sokuqeqeshwa kwemodeli, i-AI ingabheka iziqondiso zonjiniyela, izinkomba ze-API, kanye nezibonelo ezinconywayo ngesikhathi sangempela. Lokhu kunciphisa ingozi yokukhiqiza izingcingo ezindleleni noma amaphethini aphelelwe yisikhathi angahambisani nezinguqulo zakamuva ze-iOS noma i-macOS—inkinga evamile lapho kuhlelo ngaphandle kwalolu xhumano.
Emongweni waseYurophu, lapho izitudiyo ezincane ezigcina izinhlelo zokusebenza eziningana ngasikhathi sinye zivame khona, lokhu kuhlanganiswa kwe ukuzenzekela kanye nokuhambisana nemikhuba emihle ye-Apple Kungenza umehluko phakathi kwephrojekthi esala ngemuva ngokuhambisana kanye nephrojekthi evumelana ngokushesha nenguqulo ngayinye entsha yesistimu yokusebenza.
Indlela ama-ejenti asebenza ngayo ngaphakathi kwe-Xcode: ukugeleza, ukucaca, kanye nokubuyiselwa emuva
Ukuziphatha kwama-ejenti ku-Xcode 26.3 kuhluka kuphethini yakudala "yemibuzo nezimpendulo" yengxoxo ejwayelekile. Lapha, lapho ebuza umsebenzi, i-ejenti hlukanisa umgomo ube yizinyathelo ezincane okuyinto i-IDE ngokwayo eyiqopha futhi ibonise ngokulandelana, ukuze unjiniyela akwazi ukulandela intambo ngaso sonke isikhathi.
Uma uqala iseshini, i-ejenti ivame ukuthi:
- Hlaziya isakhiwo sephrojekthi bese ukhombisa ukuthi yiziphi izingxenye ezizothinta.
- Phakamisa isu lokufeza umgomo (isibonelo, ukuthi yimaphi amafayela okufanele uwadale, yiziphi izivivinyo okufanele uzengeze nokuthi ungazifaka kanjani).
- Sebenzisa izinguquko zekhodi ngokuzigqamisa ngokubonakalayo kumhleli we-Xcode.
- Qoqa iphrojekthi bese uqhuba izivivinyo, uqoqe imiphumela kusuka kukhonsoli kanye nokubuka kuqala.
- Phinda umjikelezo uma kwenzeka amaphutha, ulungise uhlelo lwakho uze uthole isakhiwo esizinzile.
Kuyo yonke le nqubo, i-Xcode idala izigigaba ezizenzakalelayo (izifinyezo) Njalo uma umenzeli eshintsha iphrojekthi, uma kukhona okungahambelani—ngoba indlela ayihambisani nesitayela seqembu, ngoba ingxenye ebalulekile iyaphuka, noma ngoba nje umphumela awugculisi—unjiniyela angabuyela kunoma yisiphi isithombe sangaphambilini ngokuchofoza kanye bese elahla lonke ibhulokhi eliyinkinga lezinguquko.
Le ndlela ifanelana kahle nezidingo zezinkampani zaseYurophu ezisebenza emikhakheni elawulwayo noma ezinezinkontileka ezidinga kakhulu. I-AI ingathatha ingxenye enkulu yemithwalo enzima, kodwa ukulandeleka kokuguqulwa ngakunye Futhi ikhono lokulungisa zonke izimo zezinguquko liyisisekelo sokungafaki engcupheni ukuzinza kwemikhiqizo esivele ikhiqizwa.
Okubaluleke kakhulu indima yokufundisa yalolu hlelo: ukubona ukuthi umenzeli uchaza kanjani izinguquko, ukuthi yimaphi amadokhumenti abonisana nawo, nokuthi yiziphi izinqumo azenzayo ezinikeza ifasitela lokufunda okusebenzayo konjiniyela abancane noma amaphrofayili asetshenziswa kabusha ohlelweni lwe-Apple.
I-Model Context Protocol (MCP): isisekelo sobuchwepheshe esivula i-Xcode ibe yi-AI eyengeziwe
Ukuhlanganiswa kwe-Xcode 26.3 nama-ejenti e-Anthropic kanye ne-OpenAI kusekelwe ku- I-Model Context Protocol (MCP), indinganiso evulekile ekuqaleni eqhutshwa yi-Anthropic echaza indlela amamodeli e-AI okufanele axhumane ngayo namathuluzi angaphandle ngendlela ephephile nehlelekile.
Uma kwenzeka i-IDE ye-Apple, i-MCP isetshenziswa Veza amakhono angaphakathi e-Xcode njengamathuluzi angabizwaUmhlanganisi, uhlelo lokuhlola, umphathi wefayela, ukubuka kuqala kwe-SwiftUI, noma amadokhumenti ngokwawo kuba imisebenzi i-ejenti engayibiza ngokulandela imithetho ecacile.
Lokhu kunemiphumela emibili ebalulekile ohlelweni lwentuthuko lwaseYurophu:
- Ngakolunye uhlangothi, i-Xcode ayikhawulelwe ku-Anthropic kanye ne-OpenAI. Noma yimuphi umenzeli ohambisana ne-MCP—kufaka phakathi lawo inkampani ewathuthukisa ngaphakathi—angaxhuma ku-IDE futhi asebenzise amakhono afanayo, okuvumela ukwakhiwa kwe izixazululo ezenzelwe imikhakha ethile (ezezimali, ezempilo, ukuphathwa komphakathi, njll.).
- Ngakolunye uhlangothi, iphrothokholi isebenza njengomngcele: ichaza lokho i-AI engakukubona nokukwenza ngaphakathi kwe-IDE, ukuze abanakho ukufinyelela okungenamkhawulo ohlelweni noma izinsiza iqembu elithanda ukuzigcina zingafinyeleleki ku-ejenti.
Ezinkampanini zokubonisana zaseSpain kanye nezitudiyo zesofthiwe, i-MCP ivula umnyango wentuthuko ama-agent akhethekile ezizenzakalela imisebenzi ethile, njengokusebenzisa iziqondiso zesitayela sangaphakathi, ukuhlola ukufinyeleleka ngokwemithetho yendawo, noma ukuqinisekisa ukuthi ikhodi iyahambisana nezimfuneko ezibekwe amaklayenti aseYurophu mayelana nokuphepha nokuvikelwa kwedatha.
Esimweni lapho imithetho yedijithali ye-EU ithola isisindo — ngezinhlaka ezifana ne-Digital Services Regulation noma i-AI Regulation eseduzane — ukukwazi ukuvumelanisa ukugeleza kwentuthuko nezidingo ezithile zomthetho ngaphandle kokunciphisa ijubane le-AI kungaba yinzuzo ecacile yokuncintisana.
Izindleko, ukumiswa kanye nokucatshangelwa kwemishini eSpain naseYurophu
Ngokombono ongokoqobo, ukusebenzisa i-Anthropic kanye ne-OpenAI AI ku-Xcode 26.3 akuyona inkimbinkimbi kakhulu, kodwa kubalulekile ukuqaphela eminye imiphumela.
Uhlobo lwe I-Xcode 26.3 esigabeni sokukhishwa kwe-Candidate Itholakala kumalungu ohlelo lonjiniyela be-Apple ngewebhusayithi esemthethweni, futhi ukuqaliswa kwayo okuzinzile kuzonwetshwa nge-App Store kamuva. Uma isifakiwe, isigaba sezilungiselelo sinikezelwe ku Ama-ejenti nabahlinzeki be-AI Ikuvumela ukuthi ulande i-Claude Agent kanye ne-Codex, uxhumanisa ngayinye ne-akhawunti ehambisanayo ngokungena ngemvume noma ukhiye we-API.
I-Anthropic kanye ne-OpenAI zisebenza kuzo zombili izimo a imodeli yemali yokusetshenziswa esuselwe kumathokheni kuthunyelwe futhi kwamukelwe. Lokhu kusho ukuthi uma iqembu lijwayela ukudlulisela ukuhlelwa kabusha okuphelele, ukukhiqiza amavolumu amakhulu ekhodi, noma ukuhlaziywa kwephrojekthi okubanzi ku-AI, ukusetshenziswa kungakhuphuka kakhulu uma kungaqashwa. Ezindaweni zaseYurophu, lapho izabelomali ze-IT zivame ukuqashelwa eduze, indlela efanelekile yile:
- Qala ngo izimo zokusetshenziswa okulinganiselwe, njengencazelo yekhodi yakudala noma ukukhiqizwa kokuhlolwa.
- Lungiselela imikhawulo yokusebenzisa imali kanye nezaziso kuma-akhawunti e-Anthropic kanye ne-OpenAI.
- Chaza ngaphakathi ukuthi yimiphi imisebenzi ehlala izenzakalelayo nokuthi yimiphi edinga ukubuyekezwa okuhlelekile ngesandla.
Esinye isici esibalulekile ezinkampanini eSpain nase-EU yi- ukuvikelwa kwempahla yobuhlakani kanye nobumfihloNakuba ikhodi yomthombo ingabhekwa njengedatha yomuntu siqu, izinhlangano eziningi zikhetha ukwazi ukuthi icutshungulwa kuphi nokuthi isetshenziswa kanjani. Bobabili i-Anthropic ne-OpenAI bagcizelele ukuzibophezela kwabo ekungasebenzisi ikhodi ethunyelwe ngalezi zihlanganisi ukuqeqesha amamodeli ngaphandle kwemvume ecacile, kodwa abaphathi bezobuchwepheshe kanye nezomthetho kufanele babuyekeze ngokucophelela le migomo, ikakhulukazi lapho besebenza namakhasimende emkhakheni womphakathi noma ezindaweni ezilawulwa kakhulu.
Kulabo abafuna ukunciphisa ukuvezwa, i-Xcode 26.3 ikuvumela ukuthi ukhawulele ukuthi yiziphi izingxenye zephrojekthi ezabelwana ngazo nama-ejenti noma uhlanganise ukusetshenziswa kwe-AI esekelwe efwini namasu okulondoloza kakhulu kumamojula abucayi kakhulu, ngaleyo ndlela kugcinwe ibhalansi phakathi isivinini sentuthuko kanye nokulawulwa kwezingozi.
Sekukonke, i-Xcode 26.3 ibeka ama-ejenti e- I-Anthropic ne-OpenAI Kusenhliziyweni yentuthuko [engacacile], kodwa ikwenza ngendlela ezama ukulinganisela ukuzimela nokuphepha: I-AI ingadala, iguqule, ihlanganise, futhi ihlole amaphrojekthi cishe njengelinye ilungu leqembu, kuyilapho unjiniyela waseYurophu egcina ukubonakala okugcwele, ikhono lokubuyela emuva ngokushesha, kanye nokuguquguquka kokuvumelanisa ukuhlanganiswa nezidingo zabo zomthetho, ezomnotho, nezobuchwepheshe. Kuma-studio kanye nama-startup eSpain, lokhu kuhlanganiswa kokuzenzakalela okujulile kanye nokulawula okulungisiwe kahle kwenza inguqulo entsha ye-Xcode ibe ithuluzi okufanele icatshangelwe ngokujulile emijikelezweni yomkhiqizo ezayo.
