# Keepcon MCP for AI Agents AI Agent Connect > Keepcon MCP for AI Agents. Automate your content moderation and semantic analysis. It lets your agent handle real-time text filtering, batch processing, and user profile lookups. Use it to keep your community safe without manual overhead. ## Overview - **Category:** customer-support - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_ghG8MoSw9hNGWQARPkVTWDTALw5fDoL7iD8Q6DqR/ai-agent-connect - **Tags:** content-moderation, semantic-analysis, user-generated-content, automated-tagging, customer-experience ## Description Keepcon is an Connector that helps you manage the heavy lifting of content moderation. Imagine you're running a massive community forum, a high-traffic Discord community, or a live chat app. The sheer volume of user-generated content is overwhelming. You need a way to catch the bad stuff—hate speech, spam, or policy violations—before they go live and damage your brand. This Connector plugs your agent directly into the Keepcon semantic moderation engine. Instead of you manually scanning all posts at 2 AM, your agent can handle the work. It looks at the context of the text, decides if it's okay, and tags it appropriately based on your rules. You can send single messages for instant approval or dump thousands of comments into a queue for processing. It even handles the boring administrative tasks like acknowledging results or exporting data for your records. If the engine makes a mistake, you can feed it back to sharpen its accuracy over time, creating a smarter filter for your specific needs. It's a way to scale your trust and safety operations without hiring a massive team to read manual tasks. By using this through the Vinkius catalog, you get a direct line to high-quality moderation without the usual integration headaches. You get a system that understands meaning rather than just looking for banned words. It gives you a way to maintain a clean, safe environment without sacrificing speed. ## Tools ### acknowledge_results Tell the system you've seen the batch results. This keeps your moderation queue organized. ### export_results Grab all the results from a batch moderation job. It's perfect for auditing or reporting. ### submit_feedback Send feedback on a specific moderation decision. This helps the semantic engine learn from its mistakes. ### get_profile_by_social_id Pull a user profile using their Twitter or Facebook ID. It links social identities to moderated content. ### get_profile Fetch a specific user's details using their Keepcon ID. Use this for deep-diving into a user's history. ### moderate_content Get an immediate approve or reject decision on text. It also adds category tags to the result. ### list_profiles See a list of all user profiles linked to your content. This helps track active users. ### search_profiles Find specific users using various filters. Use this to isolate high-risk accounts. ### import_batch Submit a huge pile of content for asynchronous processing. It returns an ID so you can track the progress. ## Prompt Examples **Prompt:** ``` Check if this comment is okay for a gaming forum: 'I'm going to find you and hurt you.' ``` **Response:** ``` I've analyzed that comment. Keepcon marked it as **REJECTED** due to 'Threats of Violence'. It doesn't meet your community guidelines. Would you like me to list the user profile associated with this post? ``` **Prompt:** ``` Can you show me the results of the last batch we processed? ``` **Response:** ``` I found the results for the recent batch. Out of 500 comments: | Status | Count | | :--- | :--- | | Approved | 485 | | Rejected | 15 | I can export the full list for you if you'd like to see the specific reasons for the rejections. ``` **Prompt:** ``` Find any users who have been flagged for aggressive behavior. ``` **Response:** ``` I've searched the profiles. I found 3 users with 'Aggressive Behavior' tags. Notable IDs include: * **user_abc** (Twitter) * **user_xyz** (Facebook) Would you like to see more details on these profiles? ``` ## Capabilities ### Moderate text in real-time Your agent can approve or reject user content instantly while also adding category tags. ### Process large content batches Submit thousands of comments at once for asynchronous review and get results in bulk. ### Export moderation results Retrieve all processed moderation data for auditing or reporting purposes. ### Submit engine feedback Train the semantic engine to be more accurate by reporting false positives. ### Query user profiles Look up and filter user profiles associated with moderated content across different networks. ## Use Cases ### Stopping forum spam instantly A forum owner wants to stop spam. They ask their agent to check every new post; the agent uses `moderate_content` to block spam immediately. ### Auditing a month of data A dev needs to audit a month of data. They ask the agent to pull all rejected posts from last week; the agent uses `export_results`. ### Finding bad actors via social IDs A safety team wants to find bad actors. They ask the agent to find users with specific social IDs; the agent uses `get_profile_by_social_id`. ### Training the engine on false positives A community manager wants to train the system. They tell the agent to mark a specific safe post as a false positive; the agent uses `submit_feedback`. ### Viewing all moderated user profiles A manager wants a bird's eye view of the community. They ask the agent to list all user profiles in the Keepcon account using `list_profiles`. ## Benefits - Handle massive volume with `import_batch` so you don't have to manually review every single comment or post. - Get instant feedback on user messages using `moderate_content` to keep your real-time chats safe. - Improve your accuracy over time by using `submit_feedback` to teach the system what a false positive looks like. - Keep your records organized by using `acknowledge_results` to clear out your pending moderation tasks. - Audit your safety logs easily with `export_results` to see exactly what was blocked and why. - Identify high-risk users quickly by using `search_profiles` to filter through your existing user base. ## How It Works The bottom line is you get automated trust and safety without the manual grunt work. 1. Enter your Keepcon API Key and Account Number in your client. 2. Tell your agent to monitor a specific content stream or queue. 3. Receive instant moderation decisions or batch results directly in your chat. ## Frequently Asked Questions **Can Keepcon MCP moderate my Discord community?** Yes, it can handle real-time text from your community to block spam or harassment automatically. **How does Keepcon handle large volumes of content?** It uses batch processing so you can send thousands of posts at once for review without hitting real-time limits. **Can I train the moderation engine to be more accurate?** Yes, you can submit feedback on specific decisions to help the system learn what a false positive looks like for your specific rules. **Is Keepcon better than just using a list of bad words?** Yes, because it uses semantic analysis to understand the intent behind the words rather than just looking for exact matches. **Can I see who the moderated users are?** Yes, you can list and search user profiles associated with moderated content to identify high-risk accounts. **Does Keepcon work with my existing moderation queue?** It can help manage your queue by acknowledging results as you process them, keeping your workspace organized. **What is the difference between synchronic and asynchronic moderation?** Synchronic (`moderate_content`) provides an immediate decision, while asynchronic (`import_batch`) is for large volumes where results are retrieved later via the export tool. **How do I ensure results are not exported twice?** After retrieving results with `export_results`, use the `acknowledge_results` tool with the corresponding `set_id` to confirm processing. **Can I provide feedback on incorrect moderation decisions?** Yes, use the `submit_feedback` tool to report false positives or negatives, helping the Keepcon engine learn and improve over time.