[KDM-TEST-FIX-417] Fix multi-agent council timeouts and fallback by utkarsh232005 · Pull Request #417 · KDM-cli/kdm-cli

utkarsh232005 · 2026-11-13T17:52:54Z

Fixes #383.\n\n- Increase the default council watchdog to 120 seconds and support KDM_COUNCIL_TIMEOUT_MS plus --council-timeout.\n- Probe Python and Ollama before launching the council.\n- Preserve useful per-workload diagnostics through rule, cloud, or raw-summary fallback when council execution is unavailable.\n- Ensure dashboard status reflects detected errors.\n- Remove duplicate analysis option declarations introduced during the v4.0.0 merge.

coderabbitai · 2026-11-13T17:53:06Z

Caution

Review failed

The pull request is closed.

Warning

`.coderabbit.yaml` has a parsing error

The CodeRabbit configuration file in this repository has a parsing error and default settings were used instead. Please fix the error(s) in the configuration file. You can initialize chat with CodeRabbit to get help with the configuration file.

Parsing errors: 2

Validation error: Invalid input: expected string, received undefined at "reviews.path_instructions[3].path"; Invalid input: expected string, received undefined at "reviews.path_instructions[3].instructions"

Configuration instructions

Please see the configuration documentation for more information.
You can also validate your configuration using the online YAML validator.
If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: # yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json

Summary

Summary by CodeRabbit

New Features
- Multi-agent analysis now supports a configurable council timeout. Set it with --council-timeout; the default is two minutes.
Bug Fixes
- Analysis now prioritizes results with errors and provides fallback explanations when council analysis fails.
- Dashboard status now reports problems when problem items are present, even if the overall result status is OK.

The analyze command now accepts a configurable council timeout. Council execution checks dependencies and uses per-result fallback handling when execution fails. The dashboard OK status uses both problem count and result status. Safety prompt options now include an interactive setting.

Changes: Council timeout and failure handling

Layer / File(s)	Summary
Timeout configuration and resolution `src/analysis/types.ts`, `src/commands/analyze.ts`, `src/analysis/analysis.ts`, `src/agent/python-bridge.ts`	The analyze command validates `--council-timeout` as a positive safe integer and passes it through analysis options. The bridge resolves the per-call value, `KDM_COUNCIL_TIMEOUT_MS`, or the 120,000 ms default.
Council execution and fallback details `src/analysis/analysis.ts`, `src/analysis/__tests__/analysis-baseline.test.ts`, `src/__tests__/cli-flags.test.ts`, `src/__tests__/e2e-system.test.ts`, `src/ui/AnalyzeDashboard.tsx`	Council execution checks dependencies and passes the configured timeout to the bridge. It selects a result with errors when available and applies rule-engine, legacy-AI, then raw-summary fallbacks after failure. The dashboard reports OK only when there are no problem items and the status is not `ProblemDetected`. Related tests mock council dependencies and verify council diagnosis details.

Interactive prompt and UI options

Layer / File(s)	Summary
Interactive options and test rendering `src/remediation/safety-prompt.ts`, `src/__tests__/safety-prompt.test.ts`, `src/__tests__/analyze-dashboard-events.test.tsx`	`PromptSafetyOptions` adds an optional `interactive` value, which `resolveIOOptions` returns. Ink tests set interactive mode explicitly, except the null-tool case, which sets it to false.

Estimated code review effort: 4 (Complex) | ~40 minutes

Change: Bug fix · Severity of issue fixed: Medium

Sequence Diagram(s)

sequenceDiagram
  participant executeCouncilForResults
  participant checkCouncilDependencies
  participant runAgentCouncilProcess
  participant RuleEngine
  participant LegacyAI
  participant RawSummary
  executeCouncilForResults->>checkCouncilDependencies: check council availability
  executeCouncilForResults->>runAgentCouncilProcess: run council with configured timeout
  alt Council returns successfully
    runAgentCouncilProcess-->>executeCouncilForResults: return consensus
  else Dependency check or council execution fails
    executeCouncilForResults->>RuleEngine: apply fallback to results with errors
    executeCouncilForResults->>LegacyAI: use when the rule engine does not apply
    executeCouncilForResults->>RawSummary: use when legacy AI returns no response
  end

Merge Risk: Moderate · up to 33053

Pressing Ctrl+C during a multi-agent council run can still start fallback work, including external AI requests, and delay exit. A stalled Python import check can also block analysis indefinitely before the council timeout applies. Both should be fixed before merging; the failures are bounded and have simple fixes.

- [ ] Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

Share

Copilot