#verification

Builds up to a goal in a field you have no expertise in: splits it into the foundations that goal actually needs, mobilizes what you already hold, learns what is missing, and leaves every solved path behind for the next goal. Nothing is called done until a check passes.

"Done" means nothing — a menhera-style completion gate for Claude Code that blocks empty completion claims until tests actually ran green and no TODOs are left behind.

Your coding agent says it works. Gopnik tries to prove it doesn't — against the code, the delivered revision, and the checks themselves.

A persistent, risk-aware engineering protocol for Hermes Agent. Software has to earn done.

Luciazero — make coding agents prove their work. Verification-first discipline for Claude Code + Codex CLI: 9-rule doctrine, 13 skills, risk-routed reviewer, fail-open hooks.

🐾 Your agent writes the code, a different model tries to break it, and only what survives ships. Cross-model review, gated merges, human-approved.

AI 코딩 비서(Claude·Codex·Gemini)가 확인도 안 하고 '다 됐어요' 하는 걸 자동으로 막아주는 품질 안전장치 — 비개발자·바이브코더를 위한 한국어 우선 하네스.

Stop your AI agent from lying about tests. Hierarchical engineering loops with machine-checkable proof gates.

Source-bound evidence and release assurance for AI-written code.

Judge-gated AI task orchestration: context recovery, real subagents, preflight reviews, repair loops and verified delivery.

Checks whether a test-count claim is actually true - in a commit message, or in what your coding agent just told you. Runs the suite and compares. pytest, vitest, jest. Pre-commit hook, Claude Code plugin, or CLI. If no claim was made, it does nothing.

Turn Claude Code into a team that has to prove it. 44 skills, 8 review agents, a validation router and a knowledge layer where nothing is claimed without a check that could have failed.

Deterministic governance and guardrails for autonomous AI coding agents. Features AST verification, evidence-based rule execution, and native MCP support for Antigravity and Claude Code.

Free local proof reports for AI-built apps and public PRs; $149 async outside proof audit.

Claude Code skills that stop an agent from reporting work as finished before it has been demonstrated on the running system.

Production-grade low-level engineering skills for AI coding agents: 193 verified, source-backed skills (C/C++/Rust/asm/kernel/networking/embedded/Zig/GPU, safety, PQC, eBPF, io_uring, secure boot, RE, design) — VFS fops, writeback, workqueues, timers, kthreads, sk_buff, TCP CC. Every claim traced: claim -> source -> section -> skill.

Local execution harness for OpenAI Codex with durable ledgers, verification gates, memory, and replay-gated skills.

Browser acceptance checks for agent-built web apps. Zero-key starter, CLI, MCP and GitHub Action; inspect explicit assertions and reports.

Neutral self-hosted registry and toolkit for Agentic Resource Discovery.

Audit long-form LLM output for factual accuracy, hallucination risk, source quality, and actionable edit suggestions.

Stop Claude from lying to you. A Claude Code skill that fact-checks every claim against real sources — code files, live docs, web search — and flags anything it can't confirm.

The trust layer for the AI agent economy — verify capability, score risk, prove reputation before every AI transaction.

Evidence-first verification for AI coding agents — prove what changed, what was tested, and what passed.

Deterministic postcondition verification for AI agents and autonomous systems using independent observed state.

Make an AI agent's 'done' mean 'here is the command output that proves it.' Deterministic, LLM-free verification for AI coding agents: sealed plans, SHA-256 hash-chained evidence, STRIPS/PDDL reachability, and a fail-closed completion gate.

An AI-powered Fact-Checking Agent built with Python, Gradio, and Google GenAI.

RCOS operator control plane for DeepSeek Harness — GoalRunner, authority gates with explicit approval, three-truth verification (execution / capability / objective satisfaction), trust ladder and next-action, task identity rebuilt from Archon runs. Side-loaded, no fork.

Go SDK for the Agent Passport System: identity, scoped delegation and signed receipts, on RFC 8785 and Ed25519.

Public research, schemas and technical evidence for Atinamos independent verification of paid AI agent services.

How to coordinate specialized AI agents with structured delegation, persistent memory, and verification gates.

Model fusion for Claude Code — one model builds, a second model (DeepSeek) judges the results. First-principles missions, a proof ledger, and a release gate that won't let your agent stop early.

Local-first agent operating system for durable AI coding: task graphs, deterministic context packs, executable delegation, verification evidence, and approval gates.

Verified, git-aware memory for autonomous software agents. Tracks what worked, what failed, what is currently true, and the evidence that proved it.

Your agent says it's done. Verity checks the evidence.

Stop your agent from saying done when it never checked. Captures real evidence and returns a PROVEN / NOT PROVEN / UNPROVABLE verdict.

Grit: an AI company in Ankara that checks its own work. Nothing leaves unchecked.

Outcome-driven long-task execution for AI agents: bounded exploration, re-planning, independent verification, and bilingual mission reports.

골만 주고 자리를 비우는 작업을 위한 Claude Code 스킬 — goal.md로 골을 잠그고 컨펌 없이 달리며, 바깥눈이 궤도 이탈을 잡는다

Checks what a coding agent says it did against what is actually in your repository. Informs you; never approves or merges.

20 production-grade skills for AI coding agents — setup, scope, discipline, code review, security, session management, governance, ops, and quality audits (eval-leakage, doc-drift). Works with Claude Code, Codex, and Cursor.