skills.rest
m16khb
m16khbCOMMUNITY
Explore all skills by this author →
GitHub

self-verify

Execute a 95-point QA gate and analyze JSON output from self-verification commands.

1|Updated Jun 4, 2026
Search Tags:#quality assurance#AI coding#agent health#QA gate#harness verification
One-click install
npx skills add https://github.com/m16khb/agent-harness --skill self-verify
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: self-verify
Source: https://github.com/m16khb/agent-harness/tree/main/skills/self-verify
Command: npx skills add https://github.com/m16khb/agent-harness --skill self-verify

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill verifies the behavior of the agent-harness across various agents, ensuring that the harness functions as intended and maintains quality control in AI coding agent workflows.

Core Features & Use Cases

  • QA Gate: Executes a 95-point QA gate to verify harness consistency across Codex, Claude Code, and GJC.
  • Inspection of Candidates: Lists and evaluates candidates for improvement in the self-verification loop.
  • Baseline Comparison: Compares and promotes verification baselines for consistent quality control.
  • Health Confirmation: Confirms CLI/MCP/native integration health through self-verification commands.

Quick Start

Run the self-verify skill with the default parameters to initiate the 95-point QA gate: ./bin/agent-harness self-verify --seed=100 --target-score=95 --json.

Frequently Asked Questions about self-verify

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify AI coding agent harness behavior across multiple agents?▼

You can verify AI coding agent harness behavior by running a 95-point QA gate that automates self-verification commands, ensuring consistent quality control across CLI, MCP, and native integrations.

How do I check CLI, MCP, and native integration health for my coding agents?▼

To check integration health, execute a series of self-verification commands that output JSON for analysis, confirming the operational status of CLI, MCP, and native integrations across your agents.

What is the best way to compare and promote verification baselines for AI coding workflows?▼

The best way to compare baselines is using an automated self-verification loop that lists and evaluates candidates, allowing you to promote a consistent baseline for quality control.

Does the 95-point QA gate work with Codex, Claude Code, and GJC?▼

Yes, the 95-point QA gate verifies harness consistency specifically across Codex, Claude Code, and GJC agents to maintain quality control in AI coding workflows.

How do I start running a QA gate to evaluate agent harness candidates?▼

Start running the QA gate by executing the self-verify command with parameters for seed, target score, and JSON output, which handles candidate evaluation and baseline comparison automatically.

Related Skills

Explore More Software Engineering Skills→
toqoztoqoz

session-harness-runner

Execute and verify session harness packages with iteration control and convergence detection.

Community
Intermediate
shhubbhshhubbh
1

canary

Monitor a live URL for console errors, performance regressions, and page failures after deploy.

Community
Advanced
useatriumuseatrium
4

harness-development

Develop and debug Rust harness-server backends normalizing CLI output into Codex App Server V2 protocol.

Official
Advanced
dawsonblockdawsonblock

cli-anything-cloudanalyzer

Compare point clouds and trajectories against reference data using CLI metrics.

Community
Intermediate