Clean scorer directory before each review to prevent stale results

Add rm -rf before mkdir -p for /tmp/strat-assess/review/ in Step 4.
In CI, reviews run sequentially in the same job, so result files from
prior reviews (e.g., RHAISTRAT-1799) persist and get picked up by
parse_results.py when scoring the next strategy (e.g., RHAISTRAT-1803).

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Eder Ignatowicz <ignatowicz@gmail.com>

Author: ederign

.claude/skills/strategy-review/SKILL.md

@@ -79,9 +79,10 @@ Launch a strat-scorer agent to produce numeric scores for the strategy. The asse
 
 Resolve the plugin root: the bootstrap script clones it to `.context/assess-strat/`. Use this path as `{PLUGIN_ROOT}`.
 
-Create the run directory:
+Create a clean run directory (removes stale result files from prior reviews in the same CI job):
 
 ```bash
+rm -rf /tmp/strat-assess/review
 mkdir -p /tmp/strat-assess/review
 ```