diff --git a/.claude/skills/ralph/SKILL.md b/.claude/skills/ralph/SKILL.md
index 5d163ed..067ba35 100644
--- a/.claude/skills/ralph/SKILL.md
+++ b/.claude/skills/ralph/SKILL.md
@@ -39,7 +39,7 @@ Take a PRD (markdown file or text) and convert it to `prd.json` in your ralph di
 
 **Each story must be completable in ONE Ralph iteration (one context window).**
 
-Ralph spawns a fresh Claude Code/OpenCode instance per iteration with no memory of previous work. If a story is too big, the LLM runs out of context before finishing and produces broken code.
+Ralph spawns a fresh Claude Code instance per iteration with no memory of previous work. If a story is too big, the LLM runs out of context before finishing and produces broken code.
 
 ### Right-sized stories:
 - Add a database column and migration
diff --git a/CLAUDE.md b/CLAUDE.md
index 7926509..3e805d0 100644
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -44,11 +44,11 @@ uv run pytest path/to/test.py  # Run specific test
 
 ## Architecture
 
+- **.env** - Secrets/API keys located at the root of the repo (git-ignored)
 - **common/** - Global configuration via pydantic-settings
   - `global_config.yaml` - Base hyperparameters and config values
   - `<name>.yaml` - Optional split configs (loaded as root key `<name>`)
   - `global_config.py` - Config class (access via `from common import global_config`)
-  - `.env` - Secrets/API keys (git-ignored)
 - **src/** - Source code (utils/)
 - **utils/llm/** - LLM inference with DSPY (`dspy_inference.py`) and LangFuse observability
 - **tests/** - pytest tests inheriting from `TestTemplate` in `test_template.py`