菜单
Use when the user asks for Lever, leverage analysis, leverage AI, self-learning leverage reviews, agent-building decisions, bottleneck diagnosis, or whether work should become a reusable agent or system.
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。
基于 SOC 职业分类
Use when the user asks about Eidos, Eidos AGI, platform status, `eidos do`, closeout, cleanup before done, docket-task orchestration, evidence verification, praxis learning, gateway routing, vault/secrets, plugins, or which Eidos AGI specialist system should handle a task. This skill tells Codex to call the installed `eidos` CLI first; MCP and plugins should only point to CLIs, while CLIs provide progressive reveal of the deeper tool graph.
Investigate current context before predicting likely futures, failures, user objections, and concrete changes. Use when Zoltar needs to research an app, plugin, architecture, workflow, plan, decision, or implementation and answer what is likely to go wrong, what evidence supports it, what complaint or failure is being prevented, and what should change today.
Answer a specific foresight decision question with researched evidence, a clear judgment, likely futures, user complaint prediction, and concrete changes. Use when Daniel asks whether an idea, implementation, plugin, architecture, or workflow is likely to be regretted, whether it is the right thing in the right place with the right authority, or what decision should be made given future effects.
Compatibility entrypoint for Zoltar foresight. Use when Daniel mentions Zoltar, asks to predict the future, asks for second-order effects, asks what he will complain about, or wants future-backed product/spec changes; route the work through researched evidence, judgment, structured output, doer/checker handoff, and self-improvement notes rather than generic prediction.
Maintain and check the Eidos marketplace when plugins must work in both Claude Code and Codex. Use when adding, refactoring, auditing, or debugging marketplace plugins, dual-host packaging, .claude-plugin and .codex-plugin manifests, root marketplace entries, source/store/cache drift, install proof, or host-neutral plugin payloads.
Challenge whether a Zoltar prediction, product idea, plugin, workflow, category framing, or architecture is overfit to existing market patterns. Use when Daniel asks for a Challenger Matrix, anti-consensus review, category redefinition, frontier user needs, non-consensus futures, Doubter voice, or whether an answer is too derivative, market-shaped, competitor-shaped, or copied from current agent/product patterns.
| name | lever |
| description | Use when the user asks for Lever, leverage analysis, leverage AI, self-learning leverage reviews, agent-building decisions, bottleneck diagnosis, or whether work should become a reusable agent or system. |
Use the live Lever CLI for leverage framing, durable learning, and agent-build decisions.
Prefer lever on PATH. If unavailable in this local environment, use:
/Users/dshanklin/repos-personal/lever/.venv/bin/lever doctor
/Users/dshanklin/repos-personal/lever/.venv/bin/lever manifesto
/Users/dshanklin/repos-personal/lever/.venv/bin/lever score "turn repeated triage into a bounded agent" --leverage-type automation --constraint "manual routing consumes founder attention" --compounding 5 --reversibility 4 --control 4 --optionality 4
/Users/dshanklin/repos-personal/lever/.venv/bin/lever agent-build "recurring inbox triage" --role "Inbox leverage scout" --repeatability 5 --boundedness 4 --proof-loop 5 --memory-fit 4 --refusal-boundary 4 --approval-risk medium
/Users/dshanklin/repos-personal/lever/.venv/bin/lever review --focus "current bottleneck"
/Users/dshanklin/repos-personal/lever/.venv/bin/lever agent "describe the work" --leverage-type optionality
/Users/dshanklin/repos-personal/lever/.venv/bin/lever learn "durable correction" --source "human correction" --evidence "short evidence" --leverage-type decision --confidence high
/Users/dshanklin/repos-personal/lever/.venv/bin/lever lessons --limit 5
/Users/dshanklin/repos-personal/lever/.venv/bin/lever distill --limit 20
/Users/dshanklin/repos-personal/lever/.venv/bin/lever improve --focus "Lever itself" --limit 20
Lever is not a generic productivity assistant. It identifies the active constraint and decides whether the highest-leverage response is an answer, artifact, reusable system, specialist agent, or durable lesson.
Lever's posture is intentionally sharp: find the constraint, build the reusable thing, kill weak loops, upgrade proven loops, and record the durable lesson.
Use lever score or lever prioritize when there is a real resource-allocation
choice. It scores compounding, reversibility, control, and optionality, then
returns a kill/test/upgrade decision.
Use lever agent-build when deciding whether repeated work should become a
specialist agent. It scores repeatability, boundedness, proof loop, memory fit,
and refusal boundary, then returns do-not-build, prototype, or
build-specialist.
Treat lever learn as an explicit learning action. Use it when the human
corrects the frame or when verified outcomes show a durable leverage principle.
Do not silently store secrets, credentials, payment details, private messages,
or weak one-off inferences.
Use lever distill after several explicit lessons have accumulated. It
summarizes the learning log into operating takeaways; it does not mutate hidden
preferences or cross approval gates.
Use lever improve when the user asks to improve Lever or when distilled
lessons should become a concrete hardening backlog. Treat its output as a plan:
make a code, docs, tests, score-threshold, or task-state change before claiming
Lever improved.
Keep this plugin thin. Source-owned behavior and durable learning state belong in the canonical Lever CLI repo.