feat(agent): per-command agent schema with response shapes

[platinummonkey]

Summary

pup agent schema returns everything at once (~185 KB) and only describes what each command takes, not what it returns, so agents guess the output format. This adds lookups for a single command or domain that also show what the command returns.

Changes

• pup agent schema logs aggregate (or logs.aggregate, or a domain like logs) returns just that part, with a returns shape for each command. --search TERM finds commands by name or description.
• Agent-mode pup <cmd> --help now covers the exact command asked for, not its whole domain.
• pup security findings schema --search/--section returns only the matching fields instead of the full ~200 KB reference.
• Output with no arguments is unchanged.

Testing

• Unit, integration, clippy and fmt all pass.
• Tests check the size limits, error paths, and that the published returns shapes match real command output.

Follow-ups

• Making all commands share one output format, and returning errors as structured data (a breaking change, so a separate PR).
• The --jq note in agent-mode output still conflicts with the published jq paths.

---

🤖 Generated with Claude Code

[datadog-prod-us1-3]

Three distinct issues were found: a timestamp unit mismatch in the metrics query contract (milliseconds vs. seconds), a missing blank-input guard in the section filter logic, and incorrect skipping of parent-local option values during help resolution.

_{Open Bits AI session}

_{🤖 Bits Code Review · Commit d1e2550 · @DataDog review to ask questions}

[datadog-prod-us1-3]

Targeted schemas overstate the JSON envelope: non-JSON output formats and plain-text leaves such as agent guide emit shapes the published contract says are required.

_{Open Bits AI session}

_{🤖 Bits Code Review · Commit dc22579 · @DataDog review to ask questions}