#3247: [Bug]: Fix session_status tool validation when called with no arguments (#3226)

## Summary Fixes #3226 - `session_status` tool now accepts calls with no arguments as documented. ## Problem When LLM calls `session_status` without arguments, validation fails with "root: must be object" error, causing session to become unresponsive. ## Solution - Allow `undefined` args by defaulting to empty object - Default to current session when no sessionKey provided - Update schema to explicitly handle empty objects - Add test case for undefined args scenario ## Testing - ✅ All existing tests pass - ✅ New test added for undefined args case - ✅ Lint and build successful ## AI-Assisted This PR was created with AI assistance (Kiro CLI) and has been fully tested. **Changes**: +36 lines, -8 lines **Backward compatible**: Yes <!-- greptile_comment --> <h2>Greptile Overview</h2> <h3>Greptile Summary</h3> This PR updates the `session_status` tool to tolerate being invoked with no arguments by defaulting `args` to `{}` inside `execute`, and adjusts sessionKey resolution to fall back to the current session when none is provided. It also tightens the TypeBox schema by disallowing additional properties, and adds a Vitest case that covers the `undefined`-args call path. These changes fit into the existing agent tool pattern (`AnyAgentTool.execute` receiving untyped args and using `readStringParam`), while improving real-world robustness against LLM tool calls that omit parameters. <h3>Confidence Score: 4/5</h3> - This PR is safe to merge with low risk; changes are small and covered by a targeted test. - The behavioral change is narrowly scoped (defaulting undefined args to an empty object and choosing a sensible sessionKey fallback) and matches the documented tool contract. Schema tightening to `additionalProperties: false` is consistent with existing TypeBox usage. The only notable improvement is making the new test avoid an awkward double-cast, but it doesn’t affect runtime correctness. - src/agents/clawdbot-tools.session-status.test.ts <!-- greptile_other_comments_section --> _{(2/5) Greptile learns from your feedback when you react with thumbs up/down!} **Context used:** - Context from `dashboard` - CLAUDE.md ([source](https://app.greptile.com/review/custom-context?memory=fd949e91-5c3a-4ab5-90a1-cbe184fd6ce8)) - Context from `dashboard` - AGENTS.md ([source](https://app.greptile.com/review/custom-context?memory=0d0c8278-ef8e-4d6c-ab21-f5527e322f13)) <!-- /greptile_comment -->