Command Claude Code
Diagnose Kimi, GLM, and other live-provider compatibility failures.
Command Claude Code
Diagnose Kimi, GLM, and other live-provider compatibility failures.
Command Claude Code
Read-only repository inspection workflow.
Command Claude Code
Plan or review subagent usage so delegation stays bounded and useful.
Command Claude Code
Review or expand the verification matrix for the live harness surface.
Command Claude Code
Audit replay, regression, and rollback gates before promoting an evolution candidate.
Command Claude Code
Shape a user-facing workflow bundle across commands, skills, agents, plugins, and MCP.
Instructions file
Claude Code instructions for iLearn-Lab/EvoHarness, covering project instructions, validation and local notes.
Plugin Claude Code
Adds release-readiness, regression, and session-recap workflows plus a quality-gate MCP bundle.
MCP server Claude CodeCodexCursor +2
MCP server "quality-gate" as configured in iLearn-Lab/EvoHarness. Runs evo_harness.quality_gate_mcp_server with python.
Agent
Scout recent failures and classify whether the next fix belongs in code, validation, or harness workflow.
Agent
Review doctor, promotion, and session evidence together before a candidate is treated as ready.
Agent
Summarize archived session behavior and extract reusable lessons for future harness turns.
Command
Scan the workspace for recent regression patterns and the best next containment move.
Command
Summarize recent workspace sessions and extract the next improvement target.
Command
Review whether a candidate or workflow is truly ready to ship.
Plugin Claude Code
Adds documentation mapping, repair, onboarding polish, and a docs-gap MCP bundle.
Agent
Map visible docs and identify the next highest-value repair.
Agent
Review whether a new user can discover the main harness surface without digging through code.
Command
Map the current documentation surface and identify the next repair target.
Command
Improve the first-run experience for this harness workspace.
Plugin Claude Code
Adds evolution-focused commands, agents, and skills for trace triage, ecosystem growth, and validation review.
Agent
Map docs, prompts, and MCP resources into a concise navigation plan.
Agent
Review the harness ecosystem and recommend the next highest-leverage asset to add.
Agent
Review whether proposed harness changes have enough replay and regression protection.