feat(lsp): add directory support to lsp_diagnostics via extension param

- Add isDirectoryPath helper to lsp-client-wrapper.ts
- Create directory-diagnostics.ts with aggregateDiagnosticsForDirectory
- Update diagnostics-tool.ts with extension parameter for directory paths
- Update Atlas agent prompts to use extension param for directory diagnostics
- Add unit tests for isDirectoryPath and aggregateDiagnosticsForDirectory

Fixes #2362

src/agents/atlas/default.ts

@@ -184,7 +184,7 @@ task(
 After EVERY delegation, complete ALL of these steps — no shortcuts:
 
 #### A. Automated Verification
-1. \`lsp_diagnostics(filePath=".")\` → ZERO errors at project level
+1. 'lsp_diagnostics(filePath=".", extension=".ts")' → ZERO errors at project level (for directory paths, extension parameter is required)
 2. \`bun run build\` or \`bun run typecheck\` → exit code 0
 3. \`bun test\` → ALL tests pass
 
@@ -346,7 +346,7 @@ You are the QA gate. Subagents lie. Verify EVERYTHING.
 
 **After each delegation — BOTH automated AND manual verification are MANDATORY:**
 
-1. \`lsp_diagnostics\` at PROJECT level → ZERO errors
+1. 'lsp_diagnostics(filePath=".", extension=".ts")' at PROJECT level → ZERO errors (for directory paths, extension parameter is required)
 2. Run build command → exit 0
 3. Run test suite → ALL pass
 4. **\`Read\` EVERY changed file line by line** → logic matches requirements
@@ -390,7 +390,7 @@ You are the QA gate. Subagents lie. Verify EVERYTHING.
 - Trust subagent claims without verification
 - Use run_in_background=true for task execution
 - Send prompts under 30 lines
-- Skip project-level lsp_diagnostics after delegation
+- Skip project-level lsp_diagnostics after delegation (use 'filePath=".", extension=".ts"' for TypeScript projects)
 - Batch multiple tasks in one delegation
 - Start fresh session for failures/follow-ups - use \`resume\` instead
 

src/agents/atlas/gemini.ts

@@ -361,7 +361,7 @@ Subagents CLAIM "done" when:
 - Trust subagent claims without verification
 - Use run_in_background=true for task execution
 - Send prompts under 30 lines
-- Skip project-level lsp_diagnostics
+- Skip project-level lsp_diagnostics (use 'filePath=".", extension=".ts"' for TypeScript projects)
 - Batch multiple tasks in one delegation
 - Start fresh session for failures (use session_id)
 

src/agents/atlas/gpt.ts

@@ -55,7 +55,7 @@ Implementation tasks are the means. Final Wave approval is the goal.
   - Verification (use Bash for tests/build)
 - Parallelize independent tool calls when possible.
 - After ANY delegation, verify with your own tool calls:
-  1. \`lsp_diagnostics\` at project level
+  1. 'lsp_diagnostics(filePath=".", extension=".ts")' at project level (for directory paths, extension parameter is required)
   2. \`Bash\` for build/test commands
   3. \`Read\` for changed files
 </tool_usage_rules>
@@ -364,7 +364,7 @@ Your job is to CATCH THEM. Assume every claim is false until YOU personally veri
 - Trust subagent claims without verification
 - Use run_in_background=true for task execution
 - Send prompts under 30 lines
-- Skip project-level lsp_diagnostics
+- Skip project-level lsp_diagnostics (use 'filePath=".", extension=".ts"' for TypeScript projects)
 - Batch multiple tasks in one delegation
 - Start fresh session for failures (use session_id)