Agents and MCP

PyTorchConference2025_GithubRepos.json

@@ -5,7 +5,7 @@
     "category": "agent",
     "github_about_section": "an open source, extensible AI agent that goes beyond code suggestions - install, execute, edit, and test with any LLM",
     "homepage_link": "https://block.github.io/goose",
-    "github_topic_closest_fit": "mcp"
+    "github_topic_closest_fit": "ai-agents"
   },
   {
     "repo_name": "ray",
@@ -503,7 +503,9 @@
   {
     "repo_name": "GEAK-agent",
     "repo_link": "https://github.com/AMD-AGI/GEAK-agent",
-    "github_about_section": "It is an LLM-based AI agent, which can write correct and efficient gpu kernels automatically."
+    "category": "agent",
+    "github_about_section": "It is an LLM-based AI agent, which can write correct and efficient gpu kernels automatically.",
+    "github_topic_closest_fit": "ai-agents"
   },
   {
     "repo_name": "hhvm",
@@ -558,6 +560,7 @@
   {
     "repo_name": "letta",
     "repo_link": "https://github.com/letta-ai/letta",
+    "category": "agent",
     "github_about_section": "Letta is the platform for building stateful agents: open AI with advanced memory that can learn and self-improve over time.",
     "homepage_link": "https://docs.letta.com",
     "github_topic_closest_fit": "ai-agents"
@@ -577,8 +580,9 @@
   {
     "repo_name": "mcp-agent",
     "repo_link": "https://github.com/lastmile-ai/mcp-agent",
+    "category": "mcp",
     "github_about_section": "Build effective agents using Model Context Protocol and simple workflow patterns",
-    "github_topic_closest_fit": "ai-agents"
+    "github_topic_closest_fit": "mcp"
   },
   {
     "repo_name": "metaflow",