111–120 of 246 posts
But is it composable like cli? The main issue to be with MCP is the entire response ends up in the context window. Whereas a decent harness and agent is usually going to pipe together and filter many tools in one long command without spending all the extra tokens.
In this thread I’m seeing a couple of the typical skill arguments. Skills don’t pollute the context window and, skills are more flexible and composable. In this specific post the SQL example gets around this by being super low-level, inherently composable, and low context consumption itself. I’m not sure this works for everyone. Looking at the Linear MCP for example, they basically just want to expose an API surface…
MCP is nice, but does your agent use it without you explicitly asking it to do so? From my experience, instructions in `.md` files aren't enough. In fact, Claude seems to prefer the shell over even its own built-in tools (e.g. `read`, `edit` etc.) See: https://github.com/anthropics/claude-code/issues/19649 . It's like an old dog that needs to be tricked to learn new tricks.
Earlier quoted context omitted.
I don’t think the “just use a CLI” crowd really are assuming you’re a developer in a coding harness. All of those use cases you mentioned benefit from the agent having access to a temporary virtual machine with a set of standard CLI tools and the ability to write and execute arbitrary code. Most already do. ChatGPT has been running Python in the cloud to answer questions before we even had functional coding harnesses…
- CLI tools don't have standardized auth, nor do they have granular tool permissions. MCP does - CLI tools need to access your secrets in your .env. With MCP the model never sees the secrets and refresh tokens. - CLI tools don't have auditing and telemetry. MCP provides all this for free. - CLI tools imply full bash access, a nightmare in terms of security and very costly if you have to spin up a machine for every re…
In this thread I’m seeing a couple of the typical skill arguments. Skills don’t pollute the context window and, skills are more flexible and composable. In this specific post the SQL example gets around this by being super low-level, inherently composable, and low context consumption itself. I’m not sure this works for everyone. Looking at the Linear MCP for example, they basically just want to expose an API surface…
> Skills don’t pollute the context window Of course they do. Do you really think that there's a magical bag of holding for the potentially infinite amount of skills?
MCP is nice, but does your agent use it without you explicitly asking it to do so? From my experience, instructions in `.md` files aren't enough. In fact, Claude seems to prefer the shell over even its own built-in tools (e.g. `read`, `edit` etc.) See: https://github.com/anthropics/claude-code/issues/19649 . It's like an old dog that needs to be tricked to learn new tricks.
or you can just wire everything up in the shell itself keeping it as simple as possible with one single tool
If I understand correctly, you’re not answering my question, and you’re not talking about MCP either. Have you considered that maybe you’re in the wrong thread?
It is amazing that folks getting paid several hundred thousand dollars a year...reinvented RPC-over-HTTP/JSON. You too, fellow web developers, are smart enough to work at Anthropic. I would love to see a proper engineering post-mortem for how this happened.
They also reinvented CORBA and DCOM, as MCP now supports gRPC as optional protocol. The reason is that these "folks getting paid several hundred thousand dollars a year" are mostly youngsters without any care to learn about computing history, which is why IT is a fashion industry. See HTMX as another example, that is basically ASP.NET Ajax or JSF Ajax.
Earlier quoted context omitted.
MCP context bloat is a solved problem since at least February. Both OpenAI and Anthropic support Client-side MCP tool search, which makes MCP context loading as efficient as the Skills one (progressive discovery). Code mode is great but not needed anymore in 95% of use-cases.
I can turn off auto invocation of a skill. It uses zero tokens when in this mode. I can still invoke it by calling it by name. How can I do that with MCP servers?
> I couldn’t find a great CLI tool for interactively probing an MCP server What about mcp-inspector? It’s a nice tool, can be used interactively, can be used as a CLI. https://github.com/modelcontextprotocol/inspector
you can specify --url, --npx and include --auth $TOKEN on the command line. you can also interactively connect with /connect, and pixel peep transport details https://fast-agent.ai/mcp/mcp-inspect-transport/