{"id":10810,"plugin_id":"plugin_asdk_app_6a552795016c8191b525649af7df1a78","kind":"skill","collection_source":null,"comparison_source":null,"observed_at":"2026-09-30T22:58:02.118Z","digest":"11a10eac04f205d057559bb051bfc5d6dadc37c65d80019779a11b38050aa1b3","against":null,"payload":{"name":"fix-question","description":"Rewrite a loose or vague research question into a sharper, twin-friendly version that produces a better response from an August twin (or any human-like respondent). Standalone helper — invoke on its own to sharpen a question before sending to August, or let other skills (notably /feedback-loops) call it programmatically when a twin response comes back too vague to use. Does NOT call any August MCP tools — pure text rewrite. Triggers on: \"rewrite this question\", \"make this question better\", \"sharpen my ask\", \"fix this research question\", \"this question is vague\", \"/fix-question\".","included_files":[],"skill_md_contents":"---\nname: fix-question\ndescription: >\n  Rewrite a loose or vague research question into a sharper, twin-friendly\n  version that produces a better response from an August twin (or any\n  human-like respondent). Standalone helper — invoke on its own to sharpen\n  a question before sending to August, or let other skills (notably\n  /feedback-loops) call it programmatically when a twin response comes back\n  too vague to use. Does NOT call any August MCP tools — pure text rewrite.\n  Triggers on: \"rewrite this question\", \"make this question better\",\n  \"sharpen my ask\", \"fix this research question\", \"this question is vague\",\n  \"/fix-question\".\n---\n\n# Fix Question\n\nTake a user's research question (or any question destined for an August\ntwin) and return a sharpened version that will produce a more useful\nresponse. This skill does not call any tools — it is a pure prompt\nrewriter.\n\n---\n\n## When to use\n\n- User pastes a question and asks to make it better.\n- User is about to send a vague question to August and wants a rewrite first.\n- Another skill (e.g. `/feedback-loops`) invokes this programmatically\n  because a twin response was too thin.\n\n## When NOT to use\n\n- The question is already sharp — say so, don't rewrite for the sake of\n  rewriting.\n- User wants to actually run the question against a twin → use\n  `/august-feedback` or `/august-interview`.\n\n---\n\n## STEP 1 — Read the question and diagnose\n\nIdentify which of these weaknesses apply. A question can have several.\n\n1. **Abstract reference** — refers to the artifact without naming it (\"this\",\n   \"the design\", \"our idea\"). Twins do better with concrete anchors.\n2. **No decision framing** — doesn't say what the answer will inform, so\n   the twin can't calibrate depth or angle.\n3. **Compound** — asks two or three things in one sentence. Twins tend to\n   answer only one, usually the last.\n4. **Under-specified dimension** — \"what do you think\" style, no axis.\n   Better: what specifically — clarity, appeal, credibility, price,\n   emotion, next step?\n5. **Leading** — presupposes an answer (\"don't you think this is clearer?\").\n6. **Abstract detail-ask** — \"what\" instead of \"what specifically\" or\n   \"give me an example of\".\n\n---\n\n## STEP 2 — Apply the rewrite rules\n\nApply in order:\n\n- **(a) Name the artifact concretely.** Replace \"this\" / \"our thing\" with a\n  short concrete label (\"the new GNC PDP hero\", \"the two-line headline\n  variant B\").\n- **(b) State the decision the answer informs.** Add a clause like \"so we\n  can decide whether to ship this\" or \"to prioritize between A and B\".\n- **(c) Constrain to one dimension per question.** If the ask is\n  compound, split into 2–3 focused questions — return them as a numbered\n  list.\n- **(d) Ask for concrete detail.** Prefer \"what specifically…\" / \"give me\n  one example of…\" over bare \"what\".\n- **(e) Strip leading language.** Replace \"don't you think X is better\"\n  with \"how does X land — is it clearer, less clear, or the same?\".\n- **(f) Keep it under ~35 words per question.** Long prompts get skimmed.\n\nIf applying a rule would break the user's intent, don't. Preserve intent\nfirst; sharpen second.\n\n---\n\n## STEP 3 — Return the rewrite\n\n**Standalone invocation** (user is the caller):\n\n```\n*Original*\n> {user's original question}\n\n*Rewritten*\n> {sharpened question — or numbered list if split}\n\n*What changed*\n{one-line explanation, e.g. \"Named the artifact, split the compound ask\ninto two questions, added the decision context.\"}\n```\n\n**Programmatic invocation** (another skill is the caller — signaled by\nthe args being just the raw question with no framing, or by the invoking\nskill setting a `terse: true` style flag):\n\nReturn **only** the rewritten question(s). No preamble, no explanation,\nno bullets. If the rewrite is a single question, return one line. If it's\nsplit, return a numbered list. This lets the caller drop the string\nstraight into an `ask_question` MCP call.\n\n---\n\n## STEP 4 — Edge cases\n\n- **Question is already sharp.** Return the original unchanged, with a\n  one-line note: \"Question is already well-formed — no rewrite needed.\"\n- **Question is unanswerable by a twin.** (e.g. \"does this align with\n  Josh's brief?\"). Flag it: \"Twins can't evaluate brief-alignment — offer\n  a gut-reaction rewrite instead?\" and propose one.\n- **Question is too broad to sharpen without more context.** Ask the user\n  what artifact / decision / audience the question is about — but only in\n  standalone mode. In programmatic mode, do your best with what's given\n  and note the ambiguity in a comment.\n\n---\n\n## Output rules\n\n- Never invent context the user didn't provide. If they didn't say what\n  the artifact is, don't make up a name — ask, or leave a placeholder like\n  `{artifact name}`.\n- Never over-rewrite. If two of the six rules apply, apply two — not all\n  six.\n- In programmatic mode, keep output copy-paste-clean: no markdown, no\n  quotes around the question unless the caller expects them.\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}