Chore: Updated performance task in plan with some detailed action items.

implementation_plan.md

@@ -1760,6 +1760,11 @@ Notes: Record provider-specific nuances, API limitations, and testing fixtures.
 - Ran `nox -s integration_tests -- robot/provider_registry.robot robot/openai_provider.robot robot/google_provider.robot` to confirm `robot/provider_registry.robot:10`, `robot/openai_provider.robot:12`, and `robot/google_provider.robot:12` stay green with logs archived in `build/reports/robot/`.
 - These official runs satisfy the outstanding Stage 2.7 requirement for provider + streaming coverage and unblock the checklist update documented below.
 
+- 2025-12-11: Stage 7 performance baseline plan
+  - Established baseline approach for profiling startup paths using the existing CLI benchmark harness (`benchmarks/cli_benchmark.py:1-34`) and will wrap it in a nox benchmark session to capture `agents --help` / `--version` timings.
+  - Identified prompt/checkpoint tuning touchpoints for streaming builds: MemorySaver checkpointing defaults in `src/cleveragents/agents/graphs/plan_generation.py:129-133` and prompt templates in `src/cleveragents/agents/graphs/plan_generation.py:155-199`.
+  - Planned streaming/build latency sampling via `PlanService` provider flows to log wall-clock and token metrics before optimization (`src/cleveragents/application/services/plan_service.py:93-205`), with results to be recorded in Phase 2 Notes and linked to the Stage 7 checklist.
+  - Added Stage 7 Implementation Checklist subtasks for profiling, prompt trimming, and checkpoint tuning to keep future optimization work traceable.
 
 **Additional LangChain/LangGraph Tasks for Phase 4:**
 - Implement `CodeGenerationGraph` using LangGraph for multi-step code generation
@@ -4350,9 +4355,10 @@ If you can do all of the above by end of Day 1, you're on track!
     - [X] Review ADR-011 for accuracy
     - [X] Add any new architectural decisions
     - [ ] Performance optimization
-    - [ ] Profile workflow execution
-    - [ ] Optimize prompt templates
-    - [ ] Tune checkpoint frequency
+      - [ ] Add nox benchmark session wrapping `benchmarks/cli_benchmark.py` to capture baseline `agents --help` / `--version` timings and log results in Phase 2 Notes.
+      - [ ] Profile PlanGenerationGraph streaming/build paths with provider stubs and record wall-clock + token metrics in Phase 2 Notes (`src/cleveragents/application/services/plan_service.py:421-525`, `src/cleveragents/agents/graphs/plan_generation.py:129-189`).
+      - [ ] Optimize analyze/generate/validate prompt templates for token efficiency while maintaining coverage expectations, documenting before/after diffs in Phase 2 Notes (`src/cleveragents/agents/graphs/plan_generation.py:155-199`).
+      - [ ] Tune MemorySaver checkpoint frequency/configuration to balance latency and resilience; document chosen defaults in Phase 2 Notes (`src/cleveragents/agents/graphs/plan_generation.py:129-133`).
   - [ ] Stage 8: Async Infrastructure
     - [ ] Implement async command execution
     - [ ] Use asyncio per ADR-002