Skip to content

CodeMode Evaluation

Evaluate strategy=codemode against strategy=tool_call on identical harness workloads.

Suggested workflow

/rlm bench preset=dynamic_web_filtering mode=harness strategy=tool_call mcp=on mcp_server=codemode
/rlm bench preset=dynamic_web_filtering mode=harness strategy=codemode mcp=on mcp_server=codemode
/rlm bench compare candidate=latest baseline=previous min_reward_delta=0.00 min_completion_delta=0.00 max_steps_increase=0.50

Metrics to review

  • avg_reward
  • completion_rate
  • avg_steps
  • usage totals
  • codemode_chain_calls
  • codemode_search_calls
  • codemode_discovery_calls
  • codemode_guardrail_blocked

Promotion guidance

Keep CodeMode opt-in unless it meets your gate thresholds with no policy regressions.

For full gating details, see: