ai-generate-context is not working fine when running in a container.

Thank you for this wonderfull tool.

I'm trying to create the context project, if I run the project through a container the output is not fine.

podman run \
  -it --rm  \
  --pull newer \
  -e AI_API_KEY=$GEMINI_API_TOKEN \
  -e CONTEXT7_API_KEY=$CONTEXT7_PERSONAL_TOKEN \
  --volume ./:/code:z \
  --workdir /code \
  registry.gitlab.com/redhat/edge/ci-cd/ai-code-review:latest \
  ai-generate-context . \
    --provider gemini \
    --output /tmp/context.md
🧠 Analysis: .
🤖 AI Provider: gemini (gemini-3-pro-preview)
💾 Output: /tmp/context.md
✅ Context generation completed!
📊 Analysis: gemini (gemini-3-pro-preview)
📄 Context file: /tmp/context.md
📏 Size: 32,111 characters

👀 Preview:
   # Project Context for AI Code Review
   
   ## Project Overview
   
   []
   
   ## Technology Stack
   
   [{'type': 'text', 'text': '### Core Technologies\n- **Primary Language:** Python\n- **Framework/Runtime:** Python Standard Library (CLI/Tooling focus)\n- **Architecture Pattern:** Modular Python Package (Library/CLI structure)\n\n### Key Dependencies (for Context7 & API Understanding)\n- **Ruff** - Configured via `ruff.toml` for high-performance linting and code formatting rules.\n- **Setuptools** - Implied by `setup.cfg` for package configuration, metadata, and dependency management.\n- **GitLab CI** - Orchestrates build and test pipelines via `.gitlab-ci.yml`.\n\n### Development Tools & CI/CD\n- **Testing:** Inferred test suite (likely `pytest` or `unittest`) based on `tests/` directory and XML reports.\n- **Code Quality:** Ruff (linter/formatter) configured via `ruff.toml`.\n- **Build/Package:** Setuptools using `setup.cfg`.\n- **CI/CD:** GitLab CI - Configured via `.gitlab-ci.yml` for automated testing and validation.', 'extras': {'signature': 'ErchCrQhAXLI2nzN6Czytx5Apd0B/8megEDOU0EVDycPz5KBHwYUjclAL4htazKBgSDbW/hEzRcbu9/JyODTH3TeH1HetDL/aO0B/iEmZ1kpvLQE4ZBY5z8bANXqI0YpXiaqHGSirc/UbL10yNSPYZQahpJsPVaOihRIh6wcpPiQR+Dqpd2+6mgWPx+fIAudE8hOXG/hq06yUxdq032Z2wt2dkgSITVz35sr5EkTOb1Ei+mBSU8R89UCi8pczoyYgbAJoRGny0xmthHb+ugNh/28Ea8E6dStxfQBX5gpkkCDNjCSCwQP5mJD4qndkHaL4IfdznhVJlAjpD

But if I execute it in my local system, everything is working fine (I'm using Fedora 42).

AI_API_KEY=$GEMINI_API_TOKEN CONTEXT7_API_KEY=$CONTEXT7_PERSONAL_TOKEN uv run ai-generate-context ../kernel-qe-tools --provider gemini -o context.md
Using CPython 3.13.9 interpreter at: /usr/bin/python3
Creating virtual environment at: .venv
      Built ai-code-review-cli @ file:///home/isantana/git/redhat/ai-code-review
Installed 140 packages in 433ms
🧠 Analysis: ../kernel-qe-tools
🤖 AI Provider: gemini (gemini-3-pro-preview)
💾 Output: context.md
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1765296162.539827 4100262 alts_credentials.cc:93] ALTS creds ignored. Not running on GCP and untrusted ALTS is not enabled.
E0000 00:00:1765296162.562606 4100262 alts_credentials.cc:93] ALTS creds ignored. Not running on GCP and untrusted ALTS is not enabled.
✅ Context generation completed!
📊 Analysis: gemini (gemini-3-pro-preview)
📄 Context file: context.md
📏 Size: 10,341 characters

👀 Preview:
   # Project Context for AI Code Review
   
   ## Project Overview
   
   **Purpose:** Provides Python-based utilities and scripts to support kernel Quality Engineering (QE) workflows and testing.
   **Type:** Python utility library and CLI toolkit
   **Domain:** Operating System Kernel Testing & Quality Assurance
   **Key Dependencies:** Standard Python Library (specific external dependencies not listed in provided context)

Here is the container version

$ podman images registry.gitlab.com/redhat/edge/ci-cd/ai-code-review
REPOSITORY                                            TAG         IMAGE ID      CREATED       SIZE
registry.gitlab.com/redhat/edge/ci-cd/ai-code-review  latest      79058ab9e2f5  25 hours ago  607 MB