67,300 tokens is real pain. But in our analysis of 201 MCP servers, the variance is wild: PostgreSQL's MCP is 1 tool, 46 tokens. GitHub's official MCP is 80 tools, 20,444 tokens. The range is 440x.

That suggests the token problem is mostly in the schema descriptions, not the protocol. GitHub's schemas have 8,000-character descriptions, redundant type restating, embedded markdown, orchestration hints — stuff that doesn't help the model but eats context fast.

Postgres is lean because its schemas are lean. If GitHub's MCP server had Postgres-quality schemas, it would be a fraction of the tokens.

agent-friend fix server.json outputs a fixed version — drops ~30% on average, more on the worst offenders: https://github.com/0-co/agent-friend

CommandSpec looks interesting. The CLI angle makes sense for stable, well-documented APIs. Curious whether the generated CLIs compress better in practice than the MCP alternative.

Author here. The biggest gap I hit after removing MCPs was APIs without CLIs.

GitHub has gh, AWS has aws, but most SaaS APIs have nothing. That's why I'm building CommandSpec. The tool takes any OpenAPI spec and generates a typed Go CLI. Every endpoint becomes a command, every parameter becomes a flag. The AI discovers operations via `--help` instead of preloading every possible tool definition.

Happy to answer questions about the setup or CommandSpec.