### CLAUDE.md
VERDICT: PASS
CORRECTNESS: VALID
SPEC_COMPLIANCE: N/A
ISSUE_COMPLIANCE: N/A
BELIEF_COMPLIANCE: N/A
TEST_COVERAGE: N/A
INTEGRATION: N/A
REASONING: This is a new development documentation file detailing the project, setup instructions (`uv sync`), and a comprehensive workflow for fixing GitHub issues. It does not introduce any executable code or functional changes.
---

### ftl_project_expert/cli.py:summary
VERDICT: PASS
CORRECTNESS: VALID
SPEC_COMPLIANCE: N/A
ISSUE_COMPLIANCE: ADDRESSES
BELIEF_COMPLIANCE: N/A
TEST_COVERAGE: UNTESTED
INTEGRATION: WIRED
REASONING: 
- **Correctness**: The logic is sound. It caps the processed beliefs to a manageable maximum of 500 (`max_beliefs = 500`). For `reasons.db`, it selects active beliefs ordered by impact. For `beliefs.md`, it correctly uses a positive lookahead `(?=^### \S+)` with `re.split` and `re.MULTILINE` to cleanly isolate sections including their headers and body details, avoiding splitting on line breaks inside belief descriptions.
- **Issue Compliance**: Solves the failure on large belief networks by filtering/truncating input size, logging the exact prompt size in KB on exception, and passing specific CLI filters (`--status IN --by-impact`) to `reasons list`.
- **Integration**: Fully integrated with the prompt building logic, passing `total_count` and `belief_count` parameters.
- **Test Coverage**: No unit/integration tests exist for this CLI command, which is a project-wide baseline status rather than a regression of this PR.
---

### ftl_project_expert/prompts/summary.py:build_summary_prompt
VERDICT: PASS
CORRECTNESS: VALID
SPEC_COMPLIANCE: N/A
ISSUE_COMPLIANCE: ADDRESSES
BELIEF_COMPLIANCE: N/A
TEST_COVERAGE: UNTESTED
INTEGRATION: WIRED
REASONING:
- **Correctness**: The function signature change has been implemented safely with `total_count` defaulting to `0` to maintain backward compatibility.
- **Issue Compliance**: The prompt is updated to explicitly communicate to the model when the beliefs have been truncated (e.g., `"top by impact out of 10822 total"`), providing appropriate context for summarizing a sampled dataset.
- **Test Coverage**: No unit tests exist for prompt builders.
---

### SELF_REVIEW
LIMITATIONS: 
- Could not execute the commands locally because terminal tool execution is restricted by environment policy.
- Verified all CLI parameter usage, Python syntax, and regular expression flags statically, which are correct and robust.
---

### FEATURE_REQUESTS
- Include unit/integration tests for prompt builders and CLI commands under a `/tests` folder to ensure code changes can be validated automatically.
---
