@ruah-dev/verify
Verify
Did the agent actually finish?
The independent, deterministic referee for agent outputs. Turns acceptance criteria into machine-runnable checks. Anything that cannot be machine-checked is marked unverifiable — never silently passed.
See it
Quick start
Up and running in seconds
@ruah-dev/verify
$ ruah verify run
# VERDICT: PASS 4/4Features
016 criteria types (shell, regex, schema, no-diff, file, script)
02Parallel criterion execution with deterministic order
03Auto-detect stack (scripts -> starter criteria)
04Markdown PR comment generator for GitHub Actions
05Structured evidence capture in .ruah/evidence.json
06Zero LLM-as-judge hallucinations
6+
Check types
0
Silent drops
CI
Exit codes
Command Reference
Commands
ruah verify runExecute all criteria from .ruah/verify.jsonruah verify run --criteria <path>Execute a specific criteria spec fileruah verify initDetect stack and generate starter criteriaruah verify report --format mdGenerate GitHub PR-ready markdown summaryruah verify report --format jsonExport machine-readable verdict & evidenceEcosystem
Part of the ruah toolchain
CLI Router
@ruah-dev/cli
One command. Every agent tool.
Guard
@ruah-dev/guard
Can I trust the agent not to do damage — in any harness, with proof?
Optimizer
@ruah-dev/opt
Where did my tokens go?
Watch (Lite)
@ruah-dev/watch
Can I show someone what this session did?
Eval
@ruah-dev/eval
Same task — which executor wins, and what did it cost?
Converter
@ruah-dev/conv
How do I make this API agent-sized?
Orchestrator
@ruah-dev/orch
Parallel agents. Zero collisions.