The core function module provided by AnQi CMS, helping you quickly build an enterprise website
Functional Icon
Single document deletion is handled by the recycling station: Deleting formal documents is moving them to the recycling station instead of physical deletion, which can be restored at any time; For site-wide accidents - upgrade failure, data damage, recycling station cleared - only backup and recovery can be solved, AnQiCMS backup covers data including static files. The two have different divisions of labor, the recycling station does not replace the backup, and the backup is not used to retrieve single draft documents.
Functional Icon
Keyword library management is responsible for consolidating, grouping and grading all-site SEO keywords, and anchor text management is responsible for automatically linking these words within the site. The rhythm of coordination is to build a library first, then link it, and finally check the density: core keywords point to column pages, long-tail keywords point to specific articles, control the range of the same word to avoid excessive interlinking on the entire site to dilute the weight signal. Both functions are built-in modules in the content system, and they need to be used together to form a complete on-site layout.
Functional Icon
The articles produced in batch by the backend AI Chat and AI skills should first be saved as drafts before going through the review: The draft link with preview parameters can be internally verified without affecting online display; after manual review, it will be transferred to the pending release or directly scheduled for release. Sensitive word filtering and content review take effect upon submission; manuscripts produced by machines and those produced manually follow the same state chain, with no exemptions.
Functional Icon
AnQiCMS has three AI capabilities that manage different segments of the creative process: AI writing is responsible for expanding the theme into an article, AI automatic tagging and AI painting generate tags and images during the creation of the article, and automatic image matching for the title assigns images based on the content of the title. This article breaks down the writing, tagging, and matching segments, explaining the intervention points of each capability and the aspects that still require manual oversight.
Functional Icon
reCAPTCHA is set up but does not intercept the effect, the fastest way to troubleshoot is to check the consistency of the key and domain binding, then confirm that the front-end component is really in the submission process, and finally check whether the server validation and exit network are reachable to the verification interface. The captcha only blocks robot submissions, and the garbage content passing through still needs sensitive word filtering and content review as a safety net.
Functional Icon
The same technology stack description has different meanings for two types of readers: web builders should look at the running resources and deployment forms brought by GoLang and Iris framework, understanding memory and concurrency performance according to the official estimates; template and secondary developers should look at Iris' routing and rendering boundaries, and how GORM data layer maps to custom content models. Concluding with only one indicator can easily lead to selection bias.
Functional Icon
The interference code is a built-in anti-crawling capability of AnQiCMS: It adds noise characters to the text stream for bulk grabbing and copying, while the page rendered to real visitors by the browser is not affected, and it is not noticeable during normal reading. It belongs to the in-site protection layer along with sensitive word filtering, JWT authentication, anti-SQL injection, and XSS, but has different responsibilities: the interference code protects content assets, while the rest of the defense line protects the site itself. The scope of activation is controlled as needed to avoid a difference in the whole site生效.
Functional Icon
A measurement of AI search citations shows that ranking list pages account for about 21% of all citations, making them the most cited content form. To handle this form within the site, three backend capabilities are required: custom content models to fix the list entries, positions, and fields, a keyword library to unify the names of all entries, and anchor text to link the list pages and detailed articles. This article provides the transformation sequence and verification points.
Functional Icon
The AI Chat on the back end is the entry point, and the AI tool set covering content, SEO, operations, business, and other domains is the executable action. The skill system solidifies common tasks into reusable steps, and the three combined can be autonomously arranged and completed by the AI intelligent body. This article explains how the arrangement occurs using the content process of selecting, generating, automatic tagging with TDK, and library, and which links still need to be confirmed by humans.
Functional Icon
The site has two paths for multilingual versions: full-page translation automatically generates the complete HTML in the target language for published pages, suitable for content with consistent structure and frequent updates; field-by-field maintenance involves setting language versions for each field in the content model, suitable for marketing pages and product pages that require manual oversight. This article combines multilingual site configuration, custom fields, and multi-site management to explain the judgment methods for mixed use of the two methods according to the nature of the pages.
Functional Icon
When integrating AnQiCMS through the built-in MCP service with external AI clients like Claude and Cursor, integration does not mean giving up full site control. The tool interface converges according to the intent directory, uses token-based authentication to confirm the identity of the caller, and uses round-level operation approval to merge write operations into a single action that can be confirmed; backup, upgrade, multi-site and other full-site high-risk areas are turned off by default and need to be explicitly enabled, reducing the interface surface that can be mistakenly called.
Functional Icon
When importing thousands of documents in bulk, it is not recommended to turn on AI automatic tags, title images, and AI TDK all at once. First, let the content model and fields stabilize, and then open AI capabilities one by one in the import process and check them, making it easy to locate problems and also leave room for rollback in the draft and recycling bin. This article explains where each of these AI capabilities is suitable to be turned on.
