Most agent frameworks use structured tool calling — JSON schemas, tool_call_id tracking, provider-specific APIs. But for bash-heavy agents, the LLM can just write `$ ls -la` as plain text. You scan for the prefix, execute it, feed output back, repeat.
No schemas, no JSON parsing, no provider lock-in. Works with any model that outputs text. Multiple commands in one response with zero overhead.
Bash already gives you validation — if the command is wrong, you get a clear error message back, and the LLM self-corrects on the next turn. You don't need a schema layer on top of that.
Even for APIs, databases, and typed inputs — you can wrap them in a CLI. Then both humans and agents use the same tool. One interface, tested by both.
And for agent skills/knowledge — what does a structured skill system give you that `cat do-something.md` doesn't?
For agents that mostly run shell commands and read files — is the tool-calling ceremony actually buying us anything?
No schemas, no JSON parsing, no provider lock-in. Works with any model that outputs text. Multiple commands in one response with zero overhead.
Bash already gives you validation — if the command is wrong, you get a clear error message back, and the LLM self-corrects on the next turn. You don't need a schema layer on top of that.
Even for APIs, databases, and typed inputs — you can wrap them in a CLI. Then both humans and agents use the same tool. One interface, tested by both.
And for agent skills/knowledge — what does a structured skill system give you that `cat do-something.md` doesn't?
For agents that mostly run shell commands and read files — is the tool-calling ceremony actually buying us anything?
Curious how others are thinking about this.