Disable automatic local context file injection in Streamlit app

[DavidJBianco]

Motivation

• Prevent inadvertent disclosure of sensitive local data by stopping the app from automatically loading context.txt into st.session_state["local_context"] and forwarding it to external LLM providers.

Description

• Replace the prior file-read logic with an explicit default of st.session_state["local_context"] = "" in peak_assistant/streamlit/app.py, and add a comment explaining the security rationale so callers must provide context explicitly.

Testing

• python -m compileall peak_assistant/streamlit/app.py succeeded; pytest -q tests/integration/test_streamlit_workflow.py failed during collection due to a missing dependency (ModuleNotFoundError: No module named 'dotenv') in this environment.

---

Codex Task

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: fdbcc4dbbe

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Preserve explicit local context input path

Setting local_context to an empty string here removes the only Streamlit code path that ever populates this value, so every runner now receives empty context in normal app usage. In environments that rely on context.txt guidance (including launches using the --context compatibility option), this is a functional regression rather than just disabling automatic injection; the app should still support an explicit, user-controlled way to provide local context.

Useful? React with 👍 / 👎.

[DavidJBianco]

Duplicate of #76. Superseded by #86. Merged into dev via merge commit 39cb907.