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.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
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.まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
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.
日本語の概要は準備中です。原文の説明を表示しています。
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.
日本語の概要は準備中です。原文の説明を表示しています。
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.
日本語の概要は準備中です。原文の説明を表示しています。
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.
日本語の概要は準備中です。原文の説明を表示しています。
Facilitate interactive brainstorming sessions using diverse creative techniques and ideation methods. Use when the user says help me brainstorm or help me ideate.
日本語の概要は準備中です。原文の説明を表示しています。
Technical documentation specialist and knowledge curator. Use when the user asks to talk to Bri or requests the tech writer.
日本語の概要は準備中です。原文の説明を表示しています。