### ftl_project_expert/cli.py
VERDICT: CONCERN
CORRECTNESS: VALID
SPEC_COMPLIANCE: N/A
ISSUE_COMPLIANCE: N/A
BELIEF_COMPLIANCE: N/A
TEST_COVERAGE: UNTESTED
INTEGRATION: WIRED
REASONING: 
- **Correctness (VALID):** The implementation logic is highly robust and correct.
  - The `init` command correctly validates the presence of the `gh` CLI when `--github-repo` is specified (lines 234-239) and safely persists this inside `config.json`.
  - In `_fetch_artifacts`, checking `"key" in ref` is done first (lines 1410-1415), ensuring any Jira issues containing string-based keys are resolved against the primary Jira source before attempting numeric match fallbacks.
  - Numeric issue/PR lookups on Jira platform fallback gracefully to `github_source` if configured (lines 1416-1433), preventing crashes by wrapping each artifact fetch in a try-except block.
  - The `status` command updates safely use `config.get("github_repo")` (lines 2071-2072) to remain backward-compatible with legacy configurations.
- **Integration (WIRED):** The new configuration option `--github-repo` is fully threaded from the CLI `init` option, through the config file schema, down to the `research` command's `_fetch_artifacts` execution.
- **Test Coverage (UNTESTED):** The primary concern is that **there are absolutely zero tests in the entire project codebase** to verify any of this functionality. Unit and integration test coverage is non-existent.

---

### SELF_REVIEW
LIMITATIONS: 
- Since the project has no unit or integration tests, it was impossible to run existing test files or confirm coverage claims. However, static verification and dependency tracing of all modified call paths confirm that the runtime logic is sound.

---

### FEATURE_REQUESTS
- Implement a basic unit testing framework (e.g., using `pytest` and `responses` or `unittest.mock`) to verify command-line argument parsing and issue/PR artifact fetching behaviors.
