align-human

align-human is a skill for Claude Code, Codex from agentscope-ai/OpenJudge. It costs 100 tokens per session (3,305 once invoked), scanned A, original, Apache-2.0.

A method for checking whether an automatic evaluator agrees with human reviewers. It uses labeled examples to measure agreement, find bias, and assess whether human review can be reduced.

In plain words
What is it for?
Use it when you have a working judge and at least 50 human-labeled examples. It compares decisions, calculates agreement and error measures, groups disagreements, checks for bias, and creates a plan for reducing human review.
Why use it?
An automated judge may make systematic mistakes even when its results look consistent, so teams need evidence before using it as a production decision gate.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it when you have a working judge and at least 50 human-labeled examples. It compares decisions, calculates agreement and error measures, groups disagreements, checks for bias, and creates a plan for reducing human review.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/agentscope-ai/openjudge/03-align-human
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add agentscope-ai/OpenJudge --skill 03-align-human
Clone the repo
git clone --depth 1 https://github.com/agentscope-ai/OpenJudge

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for align-human

README.md
[![agentmods](https://agentmods.dev/badge/skills/agentscope-ai/openjudge/03-align-human/github.svg)](https://agentmods.dev/skills/agentscope-ai/openjudge/03-align-human)
Your own site
<a href="https://agentmods.dev/skills/agentscope-ai/openjudge/03-align-human"><img src="https://agentmods.dev/badge/skills/agentscope-ai/openjudge/03-align-human/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for align-human

Your own site · 80×15
<a href="https://agentmods.dev/skills/agentscope-ai/openjudge/03-align-human"><img src="https://agentmods.dev/badge/skills/agentscope-ai/openjudge/03-align-human.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 100 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,305 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 2 findings, up to medium

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • medium Excessive Agency · line 279
    Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
    Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
  • medium Rogue Agent · line 32
    Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.
    Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00100 $0.03305
Opus 5 $0.00050 $0.01653
Sonnet 5 $0.00020 $0.00661
Haiku 4.5 $0.00010 $0.00331

Measured 10d ago against content hash 62d9a6ebc171, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

align-human scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.

The scan reads SKILL.md. This mod also ships 1 executable file (scripts/calibration.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/eval_pipeline/03-align-human/SKILL.md · 294 lines

How it starts

The opening of the file, as written. The whole thing — 294 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Align Human

Measure whether your automatic judge agrees with human judgment, detect where and why they disagree, and build a roadmap to reduce human review over time.

When to Activate

  • You have a working judge/grader and 50+ human-labeled examples
  • You want to know if the judge is trustworthy enough to replace human review
  • You've noticed the judge's decisions being overturned by humans
  • You're preparing to deploy an evaluation as a production gate

Checklist

You MUST create a task for each item and complete them in order:

  1. Load paired data — match judge verdicts with human labels
  2. Measure TPR/TNR — confusion matrix + per-stratum breakdown
  3. Calculate agreement — Cohen's kappa, Gwet's AC1, systematic bias
  4. Run bias detection — 5 systematic bias checks
  5. Analyze disagreements — cluster patterns + diagnose root causes
  6. Build human-reduction roadmap — 4-phase transition plan
  7. Confirm and record — one confirmation, then write results

Fast path: run the bundled script

Don't hand-write the calibration statistics — that is exactly where subtle bugs hide. Run the bundled, tested script (scripts/calibration.py, standard library only, no OpenJudge dependency):

python scripts/calibration.py --pairs pairs.jsonl                 # one paired file, OR
python scripts/calibration.py --verdicts verdicts.jsonl --labels labels.jsonl --stratum-key difficulty

Paired rows look like {"id","judge":"pass|fail","human":"pass|fail","stratum"?} (judge/human may also be 1/0). It prints the confusion matrix, TPR/TNR/F1 with bootstrap 95% CIs, Cohen's kappa, Gwet's AC1 (auto-flags the kappa paradox), directional bias, per-stratum TPR/TNR, and the calibration gate verdict (calibrated / not_calibrated / insufficient_evidence; exit code 0 only if calibrated). --json for machine output, --self-test to verify it.

Read the full file on GitHub · 294 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 10d ago First seen · 294 lines · 100 tokens per session scan A 62d9a6ebc171

Subscribe to this mod's changes

align-human is a skill published in the GitHub repository agentscope-ai/OpenJudge (824 stars, last pushed 2d ago), licensed Apache-2.0. It adds 100 tokens to every session and 3,305 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

tinker-fine-tuning

Provides guidance for fine-tuning LLMs using the Tinker cloud training API from Thinking Machines Lab. Use when running supervised fine-tuning, reinforcement learning (GRPO/PPO), or LoRA training on cloud GPUs via Tinker's managed infrastructure instead of local compute.

synthetic-sciences/openscience · 62 tokens

llm-as-judge-evaluation

Evaluate LLM outputs using frontier models as judges. Use for pairwise model comparison, quality scoring with custom rubrics, and automated evaluation pipelines. Covers position bias mitigation, statistical significance, and generating preference data for DPO/RLHF.

synthetic-sciences/openscience · 56 tokens

langsmith-observability

LLM observability platform for tracing, evaluation, and monitoring. Use when debugging LLM applications, evaluating model outputs against datasets, monitoring production systems, or building systematic testing pipelines for AI applications.

synthetic-sciences/openscience · 45 tokens

ceo-setup

One-time onboarding for the executive/manager commitment workflow — delegation-heavy, meeting prep, decision capture, morning and evening digests. Creates a commitments project and installs two dashboard widgets. After successful setup this skill is excluded from selection until the marker file is deleted.

suyoumo/ClawProBench · 60 tokens

developer-setup

One-time onboarding for the developer workflow — installs github-workflow missions, creates the commitments workspace, registers per-repo projects, writes calibration memories. After successful setup this skill is excluded from selection until the marker file is deleted.

suyoumo/ClawProBench · 49 tokens

code-review

Paranoid architect review of code changes for bugs, security, missing tests, and undocumented assumptions. Works on local git diffs OR a GitHub pull request (e.g. owner/repo N). For PRs, can post findings as line-level review comments.

suyoumo/ClawProBench · 57 tokens