mercurialsolo/counsel-mcp

v0.2.2 Feature

This release adds 3 notable features for engineering teams evaluating rollout.

Published 4mo MCP Developer Tools

View tool

✓ No known CVEs patched

Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Summary

AI summary

Added multi-agent research features including query planning, citation verification, quality evaluation, and LLM-as-judge scoring.

Full changelog

Multi-Agent Research Patterns

This release adds Anthropic-inspired multi-agent features to the run_deep_research tool, enabling more rigorous and transparent research workflows.

New Features

Query Planning (use_planner)

Decomposes complex queries into focused sub-questions using Claude Opus
Facet-aware synthesis for comprehensive coverage

Citation Verification (verify_citations)

Post-research verification of source attribution
Reports verified, flagged, and unsupported findings
Overall citation quality score

Quality Evaluation (evaluate_quality)

Multi-dimensional quality assessment
Completeness, source diversity, recency, and depth scoring
Gap detection with follow-up query recommendations

LLM-as-Judge (judge_quality)

5-dimensional scoring: factual accuracy, citation accuracy, completeness, source quality, synthesis quality
Detailed feedback for research improvement

Usage

run_deep_research({
  query: "What are the economic impacts of AI on labor markets?",
  depth: "deep",
  use_planner: true,
  verify_citations: true,
  evaluate_quality: true,
  judge_quality: true
})

Full Changelog

https://github.com/mercurialsolo/counsel-mcp/compare/v0.2.1...v0.2.2

View diff on GitHub

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Share on X Share on Bluesky

Track mercurialsolo/counsel-mcp

Get notified when new releases ship.

About mercurialsolo/counsel-mcp

Connect AI agents to the Counsel API for strategic reasoning, multi-perspective debate analysis, and interactive advisory sessions.

All releases →