Rewrite a rough request into a sharp, structured prompt before acting on it. Use when the user says "reprompt this", "reprompt", "improve this prompt", "make this prompt better", "structure this request", or hands you a vague/underspecified ask you should sharpen first. Complements the automatic per-turn sharpener (the UserPromptSubmit hook) for explicit, heavier rewrites. Personalizes from ~/.stellar-build/profile.md when present.
中文简介正在准备中。当前显示原文说明。
查看安装方法安装前可以查看交给 AI 代理的指令内容。
Turns a rough ask into a tight working spec, then acts on it — the explicit, on-demand counterpart to the loop's automatic per-turn sharpener. Use it when a request is vague enough to be worth structuring out loud, or when the user literally asks you to "reprompt this".
Take the text the user wants sharpened (everything after "reprompt this:", or the preceding message if they just say "reprompt that").
If ~/.stellar-build/profile.md exists (the optimizer distills it from real usage),
read it and treat it as defaults — domain, typical scope, preferred stack,
tone — unless the request contradicts them. Honor STELLAR_BUILD_HOME if set.
Rewrite the request into this structure. Keep it tight; don't pad.
Goal: the outcome actually wanted (not the literal words)
Context: relevant assumptions, environment, prior decisions
Task: the smallest precise action(s) that achieve the goal
Constraints: scope limits, must-nots, non-negotiables
Success: how we'll know it's done right
If a genuine ambiguity would change the result, ask 1–3 sharp questions first (one at a time) and fill the spec from the answers. If it's clear enough, proceed directly.
Show the sharpened spec briefly, then execute on it (or hand off to the right
stellar-build skill — e.g. find-stellar-idea, nicole-pm, tyler-architect).
Don't make the user re-approve unless a clarifying answer is still pending.
UserPromptSubmit hook and needs no invocation; turn it off with
STELLAR_BUILD_NO_REPROMPT=1./optimize-skills
refreshes ~/.stellar-build/profile.md from your traces, which feeds these rewrites.还没有评价。欢迎分享你的使用感受。
无需登录即可发表。欢迎分享使用感受、提问，或说明在哪些环境下无法运行。
还没有评论。来写第一条吧。
Push the LLM to reconsider, refine, and improve its recent output. Use when user asks for deeper critique or mentions a known deeper critique method, e.g. socratic, first principles, pre-mortem, red team.
中文简介正在准备中。当前显示原文说明。
Agentic and machine-to-machine payments on Stellar. Covers x402 (HTTP 402 paid APIs via OZ Channels facilitator, fee-sponsored clients) and MPP (Machine Payments Protocol) in both Charge mode (per-request SAC) and Channel mode (off-chain commits, high-frequency). Defaults to USDC (SEP-41 SAC) on `stellar:testnet`/`stellar:pubnet` (CAIP-2). Use when selling a paid API to AI agents, building an x402 client, or designing a payment-channel architecture for high-frequency agent traffic.
中文简介正在准备中。当前显示原文说明。
Stellar Assets (classic) + trustlines + Stellar Asset Contract (SAC) bridge to smart contracts. Covers asset issuance, distribution, authorization flags, clawback, regulated assets, trustline management, and the SAC interop layer that exposes classic assets as SEP-41 contract tokens. Use when tokenizing real-world assets, issuing stablecoins, managing trustlines, or bridging classic assets to smart contracts.
中文简介正在准备中。当前显示原文说明。
Benchmark stellar-build skills to measure quality and the lift from optimization. Use when the user says "bench my skills", "bench-skills", "benchmark skills", "did optimization help", "measure my skills", "stellar-loop bench", or wants to quantify whether /optimize-skills actually improved things. Held-out eval prompts scored by an LLM judge, reproducible with fixed seeds. Fully local.
中文简介正在准备中。当前显示原文说明。
Facilitate interactive brainstorming sessions using diverse creative techniques and ideation methods. Use when the user says help me brainstorm or help me ideate.
中文简介正在准备中。当前显示原文说明。
Technical documentation specialist and knowledge curator. Use when the user asks to talk to Bri or requests the tech writer.
中文简介正在准备中。当前显示原文说明。