网有NETHAVE
AGENT SKILL

agent-codex-lab

Fail-closed ops calibration-loop controls adapted from OpenAI Codex / MIT EQuS qubit lab (measurement skills, interdependent chains, clear-vs-noisy escalate). Deny quantum hardware spend.

使用场景Skill与智能体开发
License 许可
MIT
最近更新
2026-11-09
Stars
★ 1
Forks
0

01 前言 Front Matter

name
agent-codex-lab
description
Fail-closed ops calibration-loop controls adapted from OpenAI Codex / MIT EQuS qubit lab (measurement skills, interdependent chains, clear-vs-noisy escalate). Deny quantum hardware spend.

03 SKILL.md 正文

Agent Codex Lab

Use when running multi-step release/ops characterization (store publish chains, IAP catalog calibration, overnight CI watches).

Instructions

  1. Give each routine step a measurement skill: template, prerequisites, success/fail criteria.
  2. Persist each result before the next interdependent step.
  3. Autonomous only when signals are clear; escalate on noisy/ambiguous evidence.
  4. Routine workflows may run long; novel work needs a narrow goal.
  5. Unattended runs require progress log + check-in + steer hook.
  6. Agents own routine; humans own design/interpretation/next-step planning.
  7. Deny quantum hardware / paid Ultra lab spend under the monthly cap.
python3 scripts/agent_codex_lab.py \
  --platform ops_calibration_loop \
  --loops scripts/tests/fixtures/agent_codex_lab.json \
  --cite cxl:play-release-calibration \
  --has-skill 1 --has-prerequisites 1 \
  --has-success-fail-criteria 1 --has-template 1 \
  --prerequisites-complete 1 --result-persisted 1 \
  --signal-clear 1 --routine-workflow 1 \
  --progress-log 1 --check-in-possible 1 --steer-hook 1 \
  --agent-owns-routine 1 --human-owns-design-analysis 1

Examples

  • Allow: Play privacy→review→publish→public version verify with persisted evidence.
  • Deny: “run overnight with no log”; “novel experiment with unbounded autonomy”; quantum fridge spend.

Performance Notes

  • Zero external spend path.
  • Ranker zeros qubit/press vanity metrics.

Troubleshooting

  • Exit 2: inspect skill, chain, signal, unattended keys.
  • See docs/AGENT_CODEX_LAB.md.

04 同仓库更多 Skills

IgorGanapolsky
IgorGanapolsky/Random-Timer

writing-plans

Use when you have a spec or requirements for a multi-step task, before touching code

内容创作者写作与文案
★ 1
IgorGanapolsky
IgorGanapolsky/Random-Timer

workflow-economics-lite

Escape AI pilot purgatory: require completion-loop workflows with one owner, baseline, economic metric, system-of-record integration, and audit trail. Block orphan chatbots and interest-as-ROI claims.

Skill与智能体开发
★ 1
IgorGanapolsky
IgorGanapolsky/Random-Timer

verification-before-completion

Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always

★ 1
IgorGanapolsky
IgorGanapolsky/Random-Timer

value-center-lite

Apply InfoQ Rohrer value-center / VSM five questions (agency + coherence). Use when planning work, shipping releases, or auditing agent autonomy claims. Fail closed on product-only slogans without WQTU/paywall value.

★ 1
IgorGanapolsky
IgorGanapolsky/Random-Timer

using-superpowers

Use when starting any conversation - establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions

★ 1
IgorGanapolsky
IgorGanapolsky/Random-Timer

using-git-worktrees

Use when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallback

★ 1