菜单
Execute a sprint phase-by-phase with validator-in-the-loop retry. For each phase: implement (code, deliverables, research, etc.) → AI validators review → retry on NEEDS_FIX → advance on PASS. Works with any domain (engineering, business, research).
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/traylinx/lope --skill lope-execute命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Ask every configured validator the same question and return N independent answers — one per model. No sprint framing, no phases, no verdict parsing. Use for any multi-perspective query: 'what do these 5 CLIs think about X?', 'compare how 3 models solve this problem', 'get a second opinion across models before I commit'. Works for any domain (engineering, research, writing, business).
Compare two files across every validator. Each model picks which file is better against explicit criteria (--criteria flag). Tally the picks; print the winner. Use for A/B review, before/after diff evaluation, migration decisions, bake-offs. Criteria is mandatory in the prompt — 'better' is never model-invented.
Run an Agent-Order-style council deliberation on a decision artifact. Six built-in templates: ADR, PRD, RFC, build-vs-buy, migration-plan, incident-review. The 7-stage protocol — independent positions, anonymized critique, revision, synthesis, rubric review, minority report — turns a scenario file into a structured artifact that has survived peer scrutiny. Use when the user asks 'should we adopt X?', 'review this ADR / PRD / RFC', 'build vs buy this capability?', 'plan the migration', 'run the incident review'. Pure adversarial reasoning — no source-file modification, no git mutations.
正在显示 SKILL.md
基于 SOC 职业分类
| name | lope-execute |
| description | Execute a sprint phase-by-phase with validator-in-the-loop retry. For each phase: implement (code, deliverables, research, etc.) → AI validators review → retry on NEEDS_FIX → advance on PASS. Works with any domain (engineering, business, research). |
For each phase: implement → validators review → retry on NEEDS_FIX → advance on PASS.
This skill is invoked two ways:
/lope-execute <sprint_doc> and expects you to run that docSPRINT-*.md in the working directory is usually a safe bet — confirm with the user if ambiguous).Each phase gets validated twice back-to-back:
Set LOPE_SINGLE_STAGE=1 to revert to legacy single-pass validation.
For engineering sprints, the quality stage also includes minimality audit by default. Validators should flag unnecessary dependencies, duplicate helpers, one-implementation abstractions, broad rewrites, and symptom patches, but should not fail spec-compliant work for style alone. Disable with LOPE_MINIMALITY=off; enforce material bloat with LOPE_MINIMALITY=enforce. Business/research domains stay off unless explicitly enabled.
PYTHONPATH=~/.lope python3 -m lope execute "$SPRINT_PATH"
c. If spec NEEDS_FIX: apply the spec-level fixes, re-validate. Quality pass won't run until spec passes.
d. If quality NEEDS_FIX: apply the quality-level fixes, re-validate.
e. On both PASS: mark phase complete, move to next
3. After all phases: run /lope-audit for the scorecard
Lope auto-detects the sprint's domain from the doc header:
The validator review task adapts automatically. Same /lope-execute command for every domain.
Use --run-timeout for the whole command; --timeout remains a per-provider-call ceiling. Inspect the emitted request plan before large work and prefer compact evidence. Automatic shaping is bounded; use chunking only when the forecast fits --max-calls and --max-chunks. For abandoned work, run lope jobs list and preview lope jobs reap --dry-run. Never use pkill -f, killall, or process-name matching.