$ ctx benchmark -- "fix failing test"
ContextOS benchmark

Summary
-------
Metric                Value
--------------------  -------------------------
Task                  fix failing test
Rules parsed          7
Actionable rules      5
Filtered rules        2
Relevant rules        1
Baseline middle-risk  1/1 relevant rules (100%)
ContextOS scheduled   1 high, 0 mid
Recency reminder      enabled

Top Rules
---------
Score  Rule                                                                                      Reasons
-----  ----------------------------------------------------------------------------------------  ----------
0.50   IMPORTANT: This project has a knowledge graph. ALWAYS use code-review-graph MCP tools...  imperative
0.00   AGENTS.md
0.00   Centralised AI agent instructions. Add coding guidelines, style guides, and project c...
0.00   Ruler concatenates all .md files in this directory (and subdirectories), starting wit...
0.00   Use `query_graph` pattern="tests_for" to check coverage.
