wavecli: reconcile schema with live surfaces

[darioAnongba]

Addresses P0-7 from #997 with the intentionally small reconciliation-and-ratchet approach.

Before:

• schema --all advertised wrong names/types, phantom fields, missing flags/commands, and incorrect MCP exposure
• output contracts had no per-command identifier/version or side-effect label
• registry drift could recur without CI noticing

After:

• reconciles the confirmed drift: activity cursor, OOR destination shape, MCP flags, MCP-only daemon.balance/ark.oor.newaddress, send wait flags, exit/create/dry-run/in-round parameters, and the missing rounds join entry
• every entry emits output_schema_id, output_schema_version: 1, and side_effect
• one focused unit test walks the real cobra tree and an in-memory real MCP server, comparing visible flag names/types, covered command presence, and the exact MCP tool set

Concrete consequence: agents can trust the runtime manifest for command construction and safety classification, while future drift fails the ordinary unit suite.

Merge order: this PR must merge last after all selected surface changes (#1021 through #1027), then be rebased once so the ratchet records the final combined surface. In particular, #1022 adds watch bounds and #1026 adds non-interactive input flags that the final manifest must include. It is deliberately a draft until that final rebase.

Validation on current main:

• make unit pkg=./cmd/wavecli/...
• go test ./cmd/wavecli/... -race
• make lint-changed-local
• make commitmsg-lint range="origin/main..HEAD"

[gemini-code-assist]

Code Review

This pull request enhances the CLI schema registry by introducing fields for output schema ID/version, side effects, positional arguments, and MCP-only flags. It updates several command schemas with new parameters, adds new MCP-only tools (daemon.balance and ark.oor.newaddress), and registers the ark.rounds.join method. Additionally, a new test suite (schema_parity_test.go) is introduced to ensure parity between the schema registry, the Cobra command tree, and the live MCP tools. Feedback on the PR suggests using cmd.Runnable() instead of checking cmd.RunE != nil in the parity test to more robustly identify runnable commands.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[gemini-code-assist]

Using cmd.Runnable() is more robust and idiomatic than checking cmd.RunE != nil. It correctly handles commands that might be implemented using Run instead of RunE now or in the future.

Suggested change

	return method, cmd.RunE != nil
	return method, cmd.Runnable()

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 3d76173fb9

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Align MCP parameters with the advertised schema

Marking ark.vtxos.list as an MCP tool makes the registry advertise its CLI parameters as the tool contract, but the live MCP handler accepts status_filter and min_amount_sat (and does not accept status, min_amount, fields, or ndjson). An agent using wavecli schema to invoke this newly advertised MCP method will therefore send unsupported arguments or silently get unfiltered results; the new test compares only tool names, so it will not catch this drift. Give MCP methods their own parameter schema or validate the live MCP input schema as well.

AGENTS.md reference: cmd/wavecli/waveclicommands/AGENTS.md:L105-L113

Useful? React with 👍 / 👎.