Skip to main content
Glama

discourse_list_user_posts

Retrieve user posts and replies from Discourse forums to analyze contributions or monitor activity. Supports pagination for accessing historical content.

Instructions

Get a list of user posts and replies from a Discourse instance, with the most recent first. Returns 30 posts per page by default. Use the page parameter to paginate (page 0 = offset 0, page 1 = offset 30, etc.).

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
usernameYes
pageNo

Implementation Reference

  • The async handler function that fetches user posts and replies from Discourse using the user_actions API endpoint, paginates with offset, formats post details including topic title, date, URL, category, and excerpt, and returns formatted text content or error.
      async ({ username, page }, _extra: any) => {
        try {
          const { base, client } = ctx.siteState.ensureSelectedSite();
          const offset = (page || 0) * 30;
    
          // The filter parameter 4,5 corresponds to posts and replies
          const data = (await client.get(
            `/user_actions.json?offset=${offset}&username=${encodeURIComponent(username)}&filter=4,5`
          )) as any;
    
          const userActions = data?.user_actions || [];
    
          if (userActions.length === 0) {
            return {
              content: [{
                type: "text",
                text: page && page > 0
                  ? `No more posts found for @${username} at page ${page}.`
                  : `No posts found for @${username}.`
              }]
            };
          }
    
          const posts = userActions.map((action: any) => {
            const excerpt = action.excerpt || "";
            const truncated = action.truncated ? "..." : "";
            const date = action.created_at || "";
            const topicTitle = action.title || "";
            const topicSlug = action.slug || "";
            const topicId = action.topic_id || "";
            const postNumber = action.post_number || "";
            const categoryId = action.category_id || "";
    
            const postUrl = `${base}/t/${topicSlug}/${topicId}/${postNumber}`;
    
            return [
              `**${topicTitle}**`,
              `Posted: ${date}`,
              `Topic: ${postUrl}`,
              categoryId ? `Category ID: ${categoryId}` : undefined,
              excerpt ? `\n${excerpt}${truncated}` : undefined,
            ].filter(Boolean).join("\n");
          });
    
          const totalShown = userActions.length;
          const pageInfo = page && page > 0 ? ` (page ${page})` : "";
          const header = `Showing ${totalShown} posts for @${username}${pageInfo}:\n\n`;
          const footer = totalShown === 30 ? `\n\nTo see more posts, use page ${(page || 0) + 1}.` : "";
    
          return {
            content: [{
              type: "text",
              text: header + posts.join("\n\n---\n\n") + footer
            }]
          };
        } catch (e: any) {
          return {
            content: [{
              type: "text",
              text: `Failed to get posts for ${username}: ${e?.message || String(e)}`
            }],
            isError: true
          };
        }
      }
    );
  • Zod schema for tool input: required 'username' string and optional 'page' non-negative integer for pagination.
    const schema = z.object({
      username: z.string().min(1),
      page: z.number().int().min(0).optional(),
    });
  • The server.registerTool call that defines and registers the 'discourse_list_user_posts' tool with its metadata, input schema, and handler.
    server.registerTool(
      "discourse_list_user_posts",
      {
        title: "List User Posts",
        description: "Get a list of user posts and replies from a Discourse instance, with the most recent first. Returns 30 posts per page by default. Use the page parameter to paginate (page 0 = offset 0, page 1 = offset 30, etc.).",
        inputSchema: schema.shape,
      },
      async ({ username, page }, _extra: any) => {
        try {
          const { base, client } = ctx.siteState.ensureSelectedSite();
          const offset = (page || 0) * 30;
    
          // The filter parameter 4,5 corresponds to posts and replies
          const data = (await client.get(
            `/user_actions.json?offset=${offset}&username=${encodeURIComponent(username)}&filter=4,5`
          )) as any;
    
          const userActions = data?.user_actions || [];
    
          if (userActions.length === 0) {
            return {
              content: [{
                type: "text",
                text: page && page > 0
                  ? `No more posts found for @${username} at page ${page}.`
                  : `No posts found for @${username}.`
              }]
            };
          }
    
          const posts = userActions.map((action: any) => {
            const excerpt = action.excerpt || "";
            const truncated = action.truncated ? "..." : "";
            const date = action.created_at || "";
            const topicTitle = action.title || "";
            const topicSlug = action.slug || "";
            const topicId = action.topic_id || "";
            const postNumber = action.post_number || "";
            const categoryId = action.category_id || "";
    
            const postUrl = `${base}/t/${topicSlug}/${topicId}/${postNumber}`;
    
            return [
              `**${topicTitle}**`,
              `Posted: ${date}`,
              `Topic: ${postUrl}`,
              categoryId ? `Category ID: ${categoryId}` : undefined,
              excerpt ? `\n${excerpt}${truncated}` : undefined,
            ].filter(Boolean).join("\n");
          });
    
          const totalShown = userActions.length;
          const pageInfo = page && page > 0 ? ` (page ${page})` : "";
          const header = `Showing ${totalShown} posts for @${username}${pageInfo}:\n\n`;
          const footer = totalShown === 30 ? `\n\nTo see more posts, use page ${(page || 0) + 1}.` : "";
    
          return {
            content: [{
              type: "text",
              text: header + posts.join("\n\n---\n\n") + footer
            }]
          };
        } catch (e: any) {
          return {
            content: [{
              type: "text",
              text: `Failed to get posts for ${username}: ${e?.message || String(e)}`
            }],
            isError: true
          };
        }
      }
    );
  • Top-level call to registerListUserPosts during MCP tools initialization in registerAllTools, with read-only permissions.
    registerListUserPosts(server, ctx, { allowWrites: false });
  • Helper function exported as registerListUserPosts that encapsulates the schema, tool registration, and handler for discourse_list_user_posts.
    export const registerListUserPosts: RegisterFn = (server, ctx) => {
      const schema = z.object({
        username: z.string().min(1),
        page: z.number().int().min(0).optional(),
      });
    
      server.registerTool(
        "discourse_list_user_posts",
        {
          title: "List User Posts",
          description: "Get a list of user posts and replies from a Discourse instance, with the most recent first. Returns 30 posts per page by default. Use the page parameter to paginate (page 0 = offset 0, page 1 = offset 30, etc.).",
          inputSchema: schema.shape,
        },
        async ({ username, page }, _extra: any) => {
          try {
            const { base, client } = ctx.siteState.ensureSelectedSite();
            const offset = (page || 0) * 30;
    
            // The filter parameter 4,5 corresponds to posts and replies
            const data = (await client.get(
              `/user_actions.json?offset=${offset}&username=${encodeURIComponent(username)}&filter=4,5`
            )) as any;
    
            const userActions = data?.user_actions || [];
    
            if (userActions.length === 0) {
              return {
                content: [{
                  type: "text",
                  text: page && page > 0
                    ? `No more posts found for @${username} at page ${page}.`
                    : `No posts found for @${username}.`
                }]
              };
            }
    
            const posts = userActions.map((action: any) => {
              const excerpt = action.excerpt || "";
              const truncated = action.truncated ? "..." : "";
              const date = action.created_at || "";
              const topicTitle = action.title || "";
              const topicSlug = action.slug || "";
              const topicId = action.topic_id || "";
              const postNumber = action.post_number || "";
              const categoryId = action.category_id || "";
    
              const postUrl = `${base}/t/${topicSlug}/${topicId}/${postNumber}`;
    
              return [
                `**${topicTitle}**`,
                `Posted: ${date}`,
                `Topic: ${postUrl}`,
                categoryId ? `Category ID: ${categoryId}` : undefined,
                excerpt ? `\n${excerpt}${truncated}` : undefined,
              ].filter(Boolean).join("\n");
            });
    
            const totalShown = userActions.length;
            const pageInfo = page && page > 0 ? ` (page ${page})` : "";
            const header = `Showing ${totalShown} posts for @${username}${pageInfo}:\n\n`;
            const footer = totalShown === 30 ? `\n\nTo see more posts, use page ${(page || 0) + 1}.` : "";
    
            return {
              content: [{
                type: "text",
                text: header + posts.join("\n\n---\n\n") + footer
              }]
            };
          } catch (e: any) {
            return {
              content: [{
                type: "text",
                text: `Failed to get posts for ${username}: ${e?.message || String(e)}`
              }],
              isError: true
            };
          }
        }
      );
    };

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/SamSaffron/discourse-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server