feat: add resend commands feature to CLI and update documentation

[Shubhdeep12]

---

Summary by cubic

Add resend commands to print the full CLI command tree as JSON for scripting and agent tooling. Integrated into the root CLI, with docs, help text, and tests.

• New Features
  • resend commands: outputs the entire command tree (subcommands, options, descriptions). Auto-prints JSON when stdout is piped; use global --json in interactive terminals.
  • Registered on the root program; added tests for JSON tree shape. Updated README and skills/resend-cli/SKILL.md to include the new discovery command.

^{Written for commit 2944669. Summary will update on new commits.}

[cubic-dev-ai]

No issues found across 5 files

[felipefreitag]

@Shubhdeep12 Interesting idea, I'm curious what the use case is here? The output is pretty big, it will take a chunk of context for an agent. Does it work better than drilling down --help commands? Do you have automated/agentic workflows using this?

[Shubhdeep12]

@Shubhdeep12 Interesting idea, I'm curious what the use case is here? The output is pretty big, it will take a chunk of context for an agent. Does it work better than drilling down --help commands? Do you have automated/agentic workflows using this?

yeah, good question. the goal is one structured dump for scripts/agents so you’re not chaining --help on every subcommand or pasting stale docs into a prompt. Humans should still use --help on the command they care about and that’s the right UX there.
Token-wise, it only really helps when you’d otherwise repeat a lot of help text or walk the whole tree.
For a single command, --help is smaller.

Same idea here (runtime introspection vs stuffing documentation for agents).