[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M71QK45D115BJAM989AVMHVT":3,"skill-same:01M71QK45D115BJAM989AVMHVT":114},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":23,"publishedAt":43,"linkOnly":44,"markdownContent":45,"markdownBody":46,"frontMatter":47,"refName":49,"commitSha":50,"skillDirectory":51,"forksCount":52,"repositoryHtmlUrl":53,"repositoryHomepageUrl":54,"ownerHtmlUrl":55,"ownerBlogUrl":6,"ownerDescription":56,"updatedAt":57,"install":58,"resources":71},"01M71QK45D115BJAM989AVMHVT","method-ideate",null,"Generates and ranks grounded research or analysis ideas from repo, literature, or topic context. Use when the user asks for ideas; developing one idea goes to method-brainstorm.","sajor2000\u002Fmethodsmith","sajor2000","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F126191042?v=4","01M71QG4M3561SH49E3S7J3GKT","JC Rojas ","01M71QG4N56X2P61YTJSRP95VD",6,"MIT","https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000021","资料检索与研究","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42],"ideate","method ideate","user","constraint","research","one","use","repo","for","ideas","feature","goes","idea","context","focus","literature","analysis","when","brainstorm","2026-11-10 13:02:02",false,"---\nname: method-ideate\ndescription: \"Generates and ranks grounded research or analysis ideas from repo, literature, or topic context. Use when the user asks for ideas; developing one idea goes to method-brainstorm.\"\nargument-hint: \"[feature, focus area, or constraint]\"\n---\n\n# Generate Improvement Ideas\n\n\n## Skill Value\n\n- **Problem it solves:** Idea generation is scattered when the subject, evidence source, or evaluation lens is unclear.\n- **Use when:** The user asks for ideas, directions, research opportunities, feature concepts, or surprise-me exploration.\n- **Output:** A ranked idea set with warrants and recommended next workflow, usually method-brainstorm for selected ideas.\n- **Ask only if:** Only when the subject is too ambiguous for grounded ideation.\n- **Do not do:** Do not skip directly to implementation or present ungrounded ideas as validated plans.\n- **Interaction:** Check repo\u002Fconfig\u002Fchat evidence first. Ask one decision-changing question at a time; use the current harness's blocking question UI when available, otherwise present numbered choices and wait.\n\n**Resolve the current year at runtime** (`date +%Y`) when dating ideation documents and checking recent ones.\n\n`method-ideate` answers \"What are the strongest ideas worth exploring?\"; `method-brainstorm` answers \"What exactly should one chosen idea mean?\"; `method-plan` answers \"How should it be built?\" This workflow produces a ranked ideation artifact in `docs\u002Fideation\u002F`. It does **not** produce requirements, plans, or code.\n\n## Focus Hint\n\n\u003Cfocus_hint> #$ARGUMENTS \u003C\u002Ffocus_hint>\n\nTreat any argument as optional context: a concept, path, constraint, or volume hint (`top 3`, `100 ideas`, `raise the bar`). With none, ideate open-ended.\n\n## Core Principles\n\n1. **Ground before ideating** - Scan the actual codebase or supplied context first; no abstract advice detached from it.\n2. **Generate many -> critique all -> explain survivors only** - Quality comes from explicit rejection with reasons, not optimistic ranking.\n3. **Route action into brainstorming** - `method-brainstorm` defines the selected idea for planning; do not skip from ideation to planning.\n\n**ADHD-inspired, CE-native pattern:** borrow isolated divergence and a separate critic from [ADHD](https:\u002F\u002Fgithub.com\u002FUditAkhourii\u002Fadhd) with a smaller default dispatch; do not import its full high-cost loop or reuse its prose.\n\n## Execution Flow\n\n### Phase 0: Resume and Scope\n\n#### 0.1 Check for Recent Ideation Work\n\nLook in `docs\u002Fideation\u002F` for documents from the last 30 days whose topic or path overlaps the focus (or an obvious recent open doc for an open-ended request). Issue-grounded and non-issue ideation are distinct topics.\n\nIf one is relevant, ask: continue from it, or start fresh. To continue: read it, summarize what was explored, preserve idea statuses, and update that file rather than creating a duplicate.\n\n#### 0.2 Subject-Identification Gate\n\nBefore classifying mode or dispatching anything, confirm the subject is identifiable.\n\n**Questioning principles (0.2 and 0.4):**\n\n- Ask only for what sub-agents need: an identifiable subject (0.2) and enough substance to say something specific (0.4, elsewhere modes only).\n- Never ask about solution direction, constraints, audience, tone, success criteria, or other subject characteristics; those belong to `method-brainstorm`.\n- Always keep \"Surprise me\" as a real option; greenfield ideation is allowed.\n- Stop once the subject is identifiable or delegated. More than 3 questions across 0.2 and 0.4 suggests `method-brainstorm` instead.\n\n**Issue-tracker intent (repo mode only).** Flag it only on explicit tracker phrasing (`github issues`, `open issues`, `issue themes`, `bug reports`, `what users are reporting`), not on focus hints like `bug in auth`. When flagged, read `references\u002Fissue-tracker-mode.md` and proceed to 0.3.\n\n**Identifiability.** Would a reader, seeing only this prompt, know what to ideate on? A prompt naming only a quality or placeholder (`improvements`, `quick wins`, `bugs` alone, empty) is vague; one naming a specific feature, page, flow, subsystem, document, or concrete topic is identifiable, however short. Being inside a repo does not make a vague prompt identifiable. When the call is not obvious, read `references\u002Fsubject-identification.md`.\n\n**The scope question** (only when repo scans and user language still leave no coherent subject):\n\n- **Stem:** \"What should the agent ideate about?\"\n- **Options:**\n  - \"Specify a subject the agent should ideate on\"\n  - \"Surprise me — let the agent decide what to focus on\"\n  - \"Cancel — let me rephrase\"\n\nRouting:\n\n- **Specify** → take the follow-up as the subject and re-check identifiability once. If still ambiguous, ask once more with \"Surprise me\" still offered. Narrow only *what* the subject is, never *how* to solve it.\n- **Surprise me** → mark the run as **surprise-me mode** and read `references\u002Fsurprise-me-mode.md` now; it replaces 0.3 classification and adds rules to 0.4, 0.6, Phase 1, and Phase 2.\n- **Cancel** → exit cleanly; tell the user they can rephrase and re-invoke.\n\n#### 0.3 Mode Classification\n\nClassify the **subject** (not the CWD), deciding in order:\n\n**Decision 0 — biomedical research question?** Positive on two or more clinical\u002Fbiostat terms: population (patients, ICU, sepsis, EHR cohort, claims, omics), exposure\u002Fintervention (drug, device, procedure, biomarker), outcome (mortality, readmission, AUC, time-to-event), or design (RCT, cohort, case-control, prediction model, target trial). Negative when the subject is software, naming, narrative, or personal even if healthcare-adjacent (\"redesign our patient-portal sign-up page\" is elsewhere-software). When positive, skip Decisions 1-2, say \"Treating this as a biomedical research question — about X.\", and load `references\u002Fbiomedical-frames.md`; it replaces only Phase 2 frames and the Phase 6 menu. Phase 1 runs normally.\n\n**Decision 1 — repo-grounded vs elsewhere.** Weigh prompt content first, topic-repo coherence second, CWD repo presence only as support. Code, files, modules, tests, workflows, or issue-tracker intent -> repo-grounded. Topics with no code surface (pricing, naming, narrative, brand, personal) -> elsewhere.\n\n**Decision 2 (elsewhere only) — software vs non-software.** A product, app, page, flow, feature, or service is **elsewhere-software**, even when ideas concern its copy, UX, pricing, or design. **Elsewhere-non-software** has no software surface: brand naming, narrative, personal decisions, non-digital business, physical products.\n\nState the approach in one plain sentence, using a domain word from the topic, never the internal label:\n\n- Repo-grounded: \"Treating this as a topic in this codebase — about X.\"\n- Elsewhere-software: \"Treating this as a product\u002Fsoftware topic outside this repo — about X.\"\n- Elsewhere-non-software: \"Treating this as a [naming | narrative | business | personal] topic — about X.\"\n\nDo not prescribe correction phrases; if the user disagrees, reclassify and re-run routing. If classification is genuinely ambiguous, read `references\u002Fsubject-identification.md` for the one-question confirmation.\n\n**Elsewhere modes:** read `references\u002Felsewhere-mode.md` for the 0.4 gate and Phase 1 dispatch. **Elsewhere-non-software** additionally must load `references\u002Funiversal-ideation.md` and follow it in place of Phase 2 frame dispatch and the Phase 6 menu; do not improvise those from memory. The §6.5 Proof Failure Ladder in `references\u002Fpost-ideation-workflow.md` still applies whenever a Proof save fails.\n\n#### 0.4 Context-Substance Gate (Elsewhere Modes Only)\n\nSkip in repo mode. In elsewhere modes, follow the gate in `references\u002Felsewhere-mode.md`.\n\n#### 0.5 Interpret Focus and Volume\n\nInfer the **focus context** (concept, path, constraint, or open-ended) and any **volume override**. Defaults:\n\n- compact divergent mode: 3 sub-agents x 4 ideas (~12 raw candidates), 3-5 survivors\n- explicit wide mode: 6 sub-agents x ~6-8 ideas (~36-48 raw), 5-7 survivors\n- direct mode: 1-3 warranted options, no ideation fan-out\n\nHonor clear overrides (`top 3`, `100 ideas`, `go deep`, `raise the bar`). **Tactical scope:** `polish`, `typo(s)`, `quick wins`, `small improvements`, `cleanup`, `small fixes` lower the Phase 2 ambition floor (waive the meeting test); otherwise aim for step-function ideas.\n\n#### 0.5b Divergence Gate\n\nChoose the smallest reasoning shape that fits:\n\n- **Direct** — no ideation sub-agents when the request is closed, tactical, canonical, a syntax\u002Flookup question, or a bug with a known root cause. Treat `quick`, `standard`, `canonical`, `textbook`, `just`, and similar as direct-path signals unless the user asks to go wide. Give 1-3 warranted options and continue to the Phase 6 review loop.\n- **Compact divergent (default)** — three isolated ideation sub-agents for an open-ended topic with multiple plausible directions.\n- **Wide divergent** — six frames only when the user explicitly asks for `ADHD mode`, `go wide`, `wide exploration`, or equivalent broad exploration. Never infer wide mode from a vague or high-stakes topic.\n\nState the selected path in the cost notice. Explicit wide requests override direct-path signals.\n\n#### 0.6 Cost Transparency Notice\n\nBefore Phase 1, state the path and agent count in one informational line: grounding agents plus 0\u002F3\u002F6 ideation agents (direct\u002Fcompact\u002Fwide); +1 each for issue intelligence or opted-in Slack; -1 for a web-research skip or cache reuse. Example: \"Compact divergent path: will dispatch ~6 agents: codebase scan + learnings + web research + 3 isolated ideation sub-agents. Skip phrases: 'no external research', 'no slack'.\"\n\n### Phase 1: Mode-Aware Grounding\n\nGenerate an 8-hex `\u003Crun-id>` once and create the cross-invocation scratch directory directly under `\u002Ftmp` (not `$TMPDIR`, not `.context\u002F`):\n\n```bash\nSCRATCH_DIR=\"\u002Ftmp\u002Fmethodsmith\u002Fmethod-ideate\u002F\u003Crun-id>\"\nmkdir -p \"$SCRATCH_DIR\"\necho \"$SCRATCH_DIR\"\n```\n\nUse the echoed path as `\u003Cscratch-dir>` for the cache and checkpoints; never delete it.\n\nDispatch grounding agents in parallel in the **foreground** with the platform's subagent primitive (`Agent`\u002F`Task` in Claude Code, `spawn_agent` in Codex); run sequentially if parallel dispatch is unavailable. Omit the `mode` parameter.\n\n**Repo mode:**\n\n1. **Quick context scan** — general-purpose sub-agent on the cheapest capable model (e.g., `model: \"haiku\"` in Claude Code):\n\n   > Read AGENTS.md (CLAUDE.md as fallback, then README.md), then list the top-level layout with the native glob tool (e.g., `Glob` `*` or `*\u002F*`). Return under 30 lines: project shape (language, framework, layout), notable conventions, pain points or gaps, leverage points. Stay shallow: no issues, templates, contribution guides, or deep code search.\n   >\n   > Focus hint: {focus_hint}\n\n2. **Learnings search** — `method-learnings-researcher` with a brief summary of the focus.\n3. **Web research** — below.\n4. **Issue intelligence** — only when flagged; see `references\u002Fissue-tracker-mode.md`.\n\n**Elsewhere modes:** use the dispatch in `references\u002Felsewhere-mode.md`.\n\n**Web research (V5, V15).** Always on unless the user said \"no external research\", \"skip web research\", or equivalent; then omit `method-web-researcher` and note the skip. Before every dispatch, read `references\u002Fweb-research-cache.md` for the session cache check and append. Pass the focus hint, a one-to-two-sentence planning context, and the mode; never pass codebase content.\n\n**Slack context** is opt-in: dispatch `method-slack-researcher` in parallel only when the user asks and Slack tools exist; otherwise mention availability, or give the install hint if they asked and none are reachable.\n\n**Consolidated grounding summary** (omit empty sections): **Codebase context** (repo) or **Topic context** (elsewhere); **Past learnings**; **Issue intelligence** (when present); **External context** (note \"(reused from earlier dispatch)\" on cache reuse); **Slack context**.\n\n**Failure handling:** warn and proceed; never block on grounding. Web research failure: \"External research unavailable: {reason}. Proceeding with internal grounding only.\" Note thin elsewhere context so Phase 2 compensates with broader generation.\n\n### Phase 2: Divergent Ideation\n\nGenerate the full candidate list before critiquing any idea.\n\n**Direct path:** dispatch no ideation agents. Give the 1-3 warranted options, state why the wider loop was unnecessary, then continue to the Phase 6 review loop.\n\n**Compact and wide paths:** dispatch parallel ideation sub-agents on the inherited model (do not tier down) and omit `mode`. Compact mode dispatches **exactly 3** sub-agents with **4 candidates each**. Wide mode dispatches **6** sub-agents with ~6-8 candidates each. Compact issue-tracker mode still uses exactly 3 theme-derived or padded frames. Adjust per-agent targets only for explicit volume overrides.\n\nGive each sub-agent only the grounding summary, focus hint, its frame, and its volume target. Generator branches are isolated: do not pass one branch's output or critique to another. Instruct every generator to produce raw candidates only -- no ranking, evaluation, or hedging -- to push past its first obvious ideas, and to ground every idea in the grounding summary.\n\n**Frames:** before dispatch, read `references\u002Fideation-frames.md` (frames and compact\u002Fwide selection rule).\n\n**Per-idea output contract (all frames, all modes):**\n\n- **title**\n- **summary** (2-4 sentences)\n- **warrant** (tagged): `direct:` quoted line, file, named issue, or user-supplied context; `external:` named prior art or research with source; or `reasoned:` a written-out first-principles argument\n- **why_it_matters** — links warrant to significance\n- **meeting_test** — one line on why this warrants team discussion (waived for tactical scope)\n\nWarrant is required, not optional. An idea without articulated warrant does not surface, however plausible. Below the meeting-test floor, it does not surface unless tactical scope applies. Stay within the subject's identity: expansions, new surfaces, retirements, and pivots are fair with warrant; replacing the subject (abandoning the project, becoming a different organization) is out.\n\nAfter compact or wide agents return:\n\n1. Merge and dedupe into one master list.\n2. Synthesize cross-cutting combinations across frames (3-5 additions at most in specified mode).\n3. Weight toward any focus without excluding stronger adjacent ideas.\n4. Spread across dimensions (workflow, reliability, extensibility, capabilities, knowledge, quality, leverage) when justified.\n\n**Separate critic pass:** the orchestrator, not a generator, critiques the merged list. Score groundedness, novelty, fit, and feasibility; cluster by underlying angle; list attractive-but-unsuitable traps with one concise reason each. In compact mode, explain and deepen only the top 2 non-trap survivors; do not dispatch additional deepening agents. Wide mode may present more survivors but still keeps criticism separate from generation.\n\nThen read `references\u002Fevidence-access.md` (PubMed and Paperclip MCP tools, rules) and run its literature pulse per survivor (**Look up here:** prior art, hit count and 2-3 PMIDs, labelled `abstract`, never a systematic search).\n\n**Checkpoint A (V17).** After synthesis, write `\u003Cscratch-dir>\u002Fraw-candidates.md` with the full list and sub-agent attribution. Best-effort: on write failure, warn and proceed.\n\nAfter merging and synthesis (or after the direct-path options), and before presenting survivors, load `references\u002Fpost-ideation-workflow.md` (non-optional; never before Phase 2 dispatch completes). It holds the filtering rubric, artifact template, quality bar, and the Phase 6 menu (Refine, Open and iterate in Proof, Brainstorm, Save and end); none appear here. \"Quickly\" means fewer sub-agents, not skipping this file.\n","# Generate Improvement Ideas\n\n\n## Skill Value\n\n- **Problem it solves:** Idea generation is scattered when the subject, evidence source, or evaluation lens is unclear.\n- **Use when:** The user asks for ideas, directions, research opportunities, feature concepts, or surprise-me exploration.\n- **Output:** A ranked idea set with warrants and recommended next workflow, usually method-brainstorm for selected ideas.\n- **Ask only if:** Only when the subject is too ambiguous for grounded ideation.\n- **Do not do:** Do not skip directly to implementation or present ungrounded ideas as validated plans.\n- **Interaction:** Check repo\u002Fconfig\u002Fchat evidence first. Ask one decision-changing question at a time; use the current harness's blocking question UI when available, otherwise present numbered choices and wait.\n\n**Resolve the current year at runtime** (`date +%Y`) when dating ideation documents and checking recent ones.\n\n`method-ideate` answers \"What are the strongest ideas worth exploring?\"; `method-brainstorm` answers \"What exactly should one chosen idea mean?\"; `method-plan` answers \"How should it be built?\" This workflow produces a ranked ideation artifact in `docs\u002Fideation\u002F`. It does **not** produce requirements, plans, or code.\n\n## Focus Hint\n\n\u003Cfocus_hint> #$ARGUMENTS \u003C\u002Ffocus_hint>\n\nTreat any argument as optional context: a concept, path, constraint, or volume hint (`top 3`, `100 ideas`, `raise the bar`). With none, ideate open-ended.\n\n## Core Principles\n\n1. **Ground before ideating** - Scan the actual codebase or supplied context first; no abstract advice detached from it.\n2. **Generate many -> critique all -> explain survivors only** - Quality comes from explicit rejection with reasons, not optimistic ranking.\n3. **Route action into brainstorming** - `method-brainstorm` defines the selected idea for planning; do not skip from ideation to planning.\n\n**ADHD-inspired, CE-native pattern:** borrow isolated divergence and a separate critic from [ADHD](https:\u002F\u002Fgithub.com\u002FUditAkhourii\u002Fadhd) with a smaller default dispatch; do not import its full high-cost loop or reuse its prose.\n\n## Execution Flow\n\n### Phase 0: Resume and Scope\n\n#### 0.1 Check for Recent Ideation Work\n\nLook in `docs\u002Fideation\u002F` for documents from the last 30 days whose topic or path overlaps the focus (or an obvious recent open doc for an open-ended request). Issue-grounded and non-issue ideation are distinct topics.\n\nIf one is relevant, ask: continue from it, or start fresh. To continue: read it, summarize what was explored, preserve idea statuses, and update that file rather than creating a duplicate.\n\n#### 0.2 Subject-Identification Gate\n\nBefore classifying mode or dispatching anything, confirm the subject is identifiable.\n\n**Questioning principles (0.2 and 0.4):**\n\n- Ask only for what sub-agents need: an identifiable subject (0.2) and enough substance to say something specific (0.4, elsewhere modes only).\n- Never ask about solution direction, constraints, audience, tone, success criteria, or other subject characteristics; those belong to `method-brainstorm`.\n- Always keep \"Surprise me\" as a real option; greenfield ideation is allowed.\n- Stop once the subject is identifiable or delegated. More than 3 questions across 0.2 and 0.4 suggests `method-brainstorm` instead.\n\n**Issue-tracker intent (repo mode only).** Flag it only on explicit tracker phrasing (`github issues`, `open issues`, `issue themes`, `bug reports`, `what users are reporting`), not on focus hints like `bug in auth`. When flagged, read `references\u002Fissue-tracker-mode.md` and proceed to 0.3.\n\n**Identifiability.** Would a reader, seeing only this prompt, know what to ideate on? A prompt naming only a quality or placeholder (`improvements`, `quick wins`, `bugs` alone, empty) is vague; one naming a specific feature, page, flow, subsystem, document, or concrete topic is identifiable, however short. Being inside a repo does not make a vague prompt identifiable. When the call is not obvious, read `references\u002Fsubject-identification.md`.\n\n**The scope question** (only when repo scans and user language still leave no coherent subject):\n\n- **Stem:** \"What should the agent ideate about?\"\n- **Options:**\n  - \"Specify a subject the agent should ideate on\"\n  - \"Surprise me — let the agent decide what to focus on\"\n  - \"Cancel — let me rephrase\"\n\nRouting:\n\n- **Specify** → take the follow-up as the subject and re-check identifiability once. If still ambiguous, ask once more with \"Surprise me\" still offered. Narrow only *what* the subject is, never *how* to solve it.\n- **Surprise me** → mark the run as **surprise-me mode** and read `references\u002Fsurprise-me-mode.md` now; it replaces 0.3 classification and adds rules to 0.4, 0.6, Phase 1, and Phase 2.\n- **Cancel** → exit cleanly; tell the user they can rephrase and re-invoke.\n\n#### 0.3 Mode Classification\n\nClassify the **subject** (not the CWD), deciding in order:\n\n**Decision 0 — biomedical research question?** Positive on two or more clinical\u002Fbiostat terms: population (patients, ICU, sepsis, EHR cohort, claims, omics), exposure\u002Fintervention (drug, device, procedure, biomarker), outcome (mortality, readmission, AUC, time-to-event), or design (RCT, cohort, case-control, prediction model, target trial). Negative when the subject is software, naming, narrative, or personal even if healthcare-adjacent (\"redesign our patient-portal sign-up page\" is elsewhere-software). When positive, skip Decisions 1-2, say \"Treating this as a biomedical research question — about X.\", and load `references\u002Fbiomedical-frames.md`; it replaces only Phase 2 frames and the Phase 6 menu. Phase 1 runs normally.\n\n**Decision 1 — repo-grounded vs elsewhere.** Weigh prompt content first, topic-repo coherence second, CWD repo presence only as support. Code, files, modules, tests, workflows, or issue-tracker intent -> repo-grounded. Topics with no code surface (pricing, naming, narrative, brand, personal) -> elsewhere.\n\n**Decision 2 (elsewhere only) — software vs non-software.** A product, app, page, flow, feature, or service is **elsewhere-software**, even when ideas concern its copy, UX, pricing, or design. **Elsewhere-non-software** has no software surface: brand naming, narrative, personal decisions, non-digital business, physical products.\n\nState the approach in one plain sentence, using a domain word from the topic, never the internal label:\n\n- Repo-grounded: \"Treating this as a topic in this codebase — about X.\"\n- Elsewhere-software: \"Treating this as a product\u002Fsoftware topic outside this repo — about X.\"\n- Elsewhere-non-software: \"Treating this as a [naming | narrative | business | personal] topic — about X.\"\n\nDo not prescribe correction phrases; if the user disagrees, reclassify and re-run routing. If classification is genuinely ambiguous, read `references\u002Fsubject-identification.md` for the one-question confirmation.\n\n**Elsewhere modes:** read `references\u002Felsewhere-mode.md` for the 0.4 gate and Phase 1 dispatch. **Elsewhere-non-software** additionally must load `references\u002Funiversal-ideation.md` and follow it in place of Phase 2 frame dispatch and the Phase 6 menu; do not improvise those from memory. The §6.5 Proof Failure Ladder in `references\u002Fpost-ideation-workflow.md` still applies whenever a Proof save fails.\n\n#### 0.4 Context-Substance Gate (Elsewhere Modes Only)\n\nSkip in repo mode. In elsewhere modes, follow the gate in `references\u002Felsewhere-mode.md`.\n\n#### 0.5 Interpret Focus and Volume\n\nInfer the **focus context** (concept, path, constraint, or open-ended) and any **volume override**. Defaults:\n\n- compact divergent mode: 3 sub-agents x 4 ideas (~12 raw candidates), 3-5 survivors\n- explicit wide mode: 6 sub-agents x ~6-8 ideas (~36-48 raw), 5-7 survivors\n- direct mode: 1-3 warranted options, no ideation fan-out\n\nHonor clear overrides (`top 3`, `100 ideas`, `go deep`, `raise the bar`). **Tactical scope:** `polish`, `typo(s)`, `quick wins`, `small improvements`, `cleanup`, `small fixes` lower the Phase 2 ambition floor (waive the meeting test); otherwise aim for step-function ideas.\n\n#### 0.5b Divergence Gate\n\nChoose the smallest reasoning shape that fits:\n\n- **Direct** — no ideation sub-agents when the request is closed, tactical, canonical, a syntax\u002Flookup question, or a bug with a known root cause. Treat `quick`, `standard`, `canonical`, `textbook`, `just`, and similar as direct-path signals unless the user asks to go wide. Give 1-3 warranted options and continue to the Phase 6 review loop.\n- **Compact divergent (default)** — three isolated ideation sub-agents for an open-ended topic with multiple plausible directions.\n- **Wide divergent** — six frames only when the user explicitly asks for `ADHD mode`, `go wide`, `wide exploration`, or equivalent broad exploration. Never infer wide mode from a vague or high-stakes topic.\n\nState the selected path in the cost notice. Explicit wide requests override direct-path signals.\n\n#### 0.6 Cost Transparency Notice\n\nBefore Phase 1, state the path and agent count in one informational line: grounding agents plus 0\u002F3\u002F6 ideation agents (direct\u002Fcompact\u002Fwide); +1 each for issue intelligence or opted-in Slack; -1 for a web-research skip or cache reuse. Example: \"Compact divergent path: will dispatch ~6 agents: codebase scan + learnings + web research + 3 isolated ideation sub-agents. Skip phrases: 'no external research', 'no slack'.\"\n\n### Phase 1: Mode-Aware Grounding\n\nGenerate an 8-hex `\u003Crun-id>` once and create the cross-invocation scratch directory directly under `\u002Ftmp` (not `$TMPDIR`, not `.context\u002F`):\n\n```bash\nSCRATCH_DIR=\"\u002Ftmp\u002Fmethodsmith\u002Fmethod-ideate\u002F\u003Crun-id>\"\nmkdir -p \"$SCRATCH_DIR\"\necho \"$SCRATCH_DIR\"\n```\n\nUse the echoed path as `\u003Cscratch-dir>` for the cache and checkpoints; never delete it.\n\nDispatch grounding agents in parallel in the **foreground** with the platform's subagent primitive (`Agent`\u002F`Task` in Claude Code, `spawn_agent` in Codex); run sequentially if parallel dispatch is unavailable. Omit the `mode` parameter.\n\n**Repo mode:**\n\n1. **Quick context scan** — general-purpose sub-agent on the cheapest capable model (e.g., `model: \"haiku\"` in Claude Code):\n\n   > Read AGENTS.md (CLAUDE.md as fallback, then README.md), then list the top-level layout with the native glob tool (e.g., `Glob` `*` or `*\u002F*`). Return under 30 lines: project shape (language, framework, layout), notable conventions, pain points or gaps, leverage points. Stay shallow: no issues, templates, contribution guides, or deep code search.\n   >\n   > Focus hint: {focus_hint}\n\n2. **Learnings search** — `method-learnings-researcher` with a brief summary of the focus.\n3. **Web research** — below.\n4. **Issue intelligence** — only when flagged; see `references\u002Fissue-tracker-mode.md`.\n\n**Elsewhere modes:** use the dispatch in `references\u002Felsewhere-mode.md`.\n\n**Web research (V5, V15).** Always on unless the user said \"no external research\", \"skip web research\", or equivalent; then omit `method-web-researcher` and note the skip. Before every dispatch, read `references\u002Fweb-research-cache.md` for the session cache check and append. Pass the focus hint, a one-to-two-sentence planning context, and the mode; never pass codebase content.\n\n**Slack context** is opt-in: dispatch `method-slack-researcher` in parallel only when the user asks and Slack tools exist; otherwise mention availability, or give the install hint if they asked and none are reachable.\n\n**Consolidated grounding summary** (omit empty sections): **Codebase context** (repo) or **Topic context** (elsewhere); **Past learnings**; **Issue intelligence** (when present); **External context** (note \"(reused from earlier dispatch)\" on cache reuse); **Slack context**.\n\n**Failure handling:** warn and proceed; never block on grounding. Web research failure: \"External research unavailable: {reason}. Proceeding with internal grounding only.\" Note thin elsewhere context so Phase 2 compensates with broader generation.\n\n### Phase 2: Divergent Ideation\n\nGenerate the full candidate list before critiquing any idea.\n\n**Direct path:** dispatch no ideation agents. Give the 1-3 warranted options, state why the wider loop was unnecessary, then continue to the Phase 6 review loop.\n\n**Compact and wide paths:** dispatch parallel ideation sub-agents on the inherited model (do not tier down) and omit `mode`. Compact mode dispatches **exactly 3** sub-agents with **4 candidates each**. Wide mode dispatches **6** sub-agents with ~6-8 candidates each. Compact issue-tracker mode still uses exactly 3 theme-derived or padded frames. Adjust per-agent targets only for explicit volume overrides.\n\nGive each sub-agent only the grounding summary, focus hint, its frame, and its volume target. Generator branches are isolated: do not pass one branch's output or critique to another. Instruct every generator to produce raw candidates only -- no ranking, evaluation, or hedging -- to push past its first obvious ideas, and to ground every idea in the grounding summary.\n\n**Frames:** before dispatch, read `references\u002Fideation-frames.md` (frames and compact\u002Fwide selection rule).\n\n**Per-idea output contract (all frames, all modes):**\n\n- **title**\n- **summary** (2-4 sentences)\n- **warrant** (tagged): `direct:` quoted line, file, named issue, or user-supplied context; `external:` named prior art or research with source; or `reasoned:` a written-out first-principles argument\n- **why_it_matters** — links warrant to significance\n- **meeting_test** — one line on why this warrants team discussion (waived for tactical scope)\n\nWarrant is required, not optional. An idea without articulated warrant does not surface, however plausible. Below the meeting-test floor, it does not surface unless tactical scope applies. Stay within the subject's identity: expansions, new surfaces, retirements, and pivots are fair with warrant; replacing the subject (abandoning the project, becoming a different organization) is out.\n\nAfter compact or wide agents return:\n\n1. Merge and dedupe into one master list.\n2. Synthesize cross-cutting combinations across frames (3-5 additions at most in specified mode).\n3. Weight toward any focus without excluding stronger adjacent ideas.\n4. Spread across dimensions (workflow, reliability, extensibility, capabilities, knowledge, quality, leverage) when justified.\n\n**Separate critic pass:** the orchestrator, not a generator, critiques the merged list. Score groundedness, novelty, fit, and feasibility; cluster by underlying angle; list attractive-but-unsuitable traps with one concise reason each. In compact mode, explain and deepen only the top 2 non-trap survivors; do not dispatch additional deepening agents. Wide mode may present more survivors but still keeps criticism separate from generation.\n\nThen read `references\u002Fevidence-access.md` (PubMed and Paperclip MCP tools, rules) and run its literature pulse per survivor (**Look up here:** prior art, hit count and 2-3 PMIDs, labelled `abstract`, never a systematic search).\n\n**Checkpoint A (V17).** After synthesis, write `\u003Cscratch-dir>\u002Fraw-candidates.md` with the full list and sub-agent attribution. Best-effort: on write failure, warn and proceed.\n\nAfter merging and synthesis (or after the direct-path options), and before presenting survivors, load `references\u002Fpost-ideation-workflow.md` (non-optional; never before Phase 2 dispatch completes). It holds the filtering rubric, artifact template, quality bar, and the Phase 6 menu (Refine, Open and iterate in Proof, Brainstorm, Save and end); none appear here. \"Quickly\" means fewer sub-agents, not skipping this file.\n",{"name":5,"description":7,"argument-hint":48},"[feature, focus area, or constraint]","main","893bccb980e4dce1345919e9f675f5a9dcbed364","plugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate",0,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith","https:\u002F\u002Fsajor2000.github.io\u002Fmethodsmith\u002F","https:\u002F\u002Fgithub.com\u002Fsajor2000","Critical care physician-scientist at Rush University. @Common-Longitudinal-ICU-data-Format contributor building open ICU data tools & clinical AI","2026-11-01 01:56:01",{"source":8,"skillName":5,"commands":59,"prompt":69,"zipAvailable":70},[60,63,66],{"tool":61,"command":62},"npx","npx skills add sajor2000\u002Fmethodsmith --skill method-ideate",{"tool":64,"command":65},"bunx","bunx skills add sajor2000\u002Fmethodsmith --skill method-ideate",{"tool":67,"command":68},"pnpm","pnpm dlx skills add sajor2000\u002Fmethodsmith --skill method-ideate","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002FSKILL.md",true,[72,78,82,86,90,94,98,102,106,110],{"path":73,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":76,"sourceHtmlUrl":77},"references\u002Fbiomedical-frames.md","REFERENCE","参考文档",7782,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002Freferences\u002Fbiomedical-frames.md",{"path":79,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":80,"sourceHtmlUrl":81},"references\u002Felsewhere-mode.md",1978,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002Freferences\u002Felsewhere-mode.md",{"path":83,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":84,"sourceHtmlUrl":85},"references\u002Fevidence-access.md",3675,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002Freferences\u002Fevidence-access.md",{"path":87,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":88,"sourceHtmlUrl":89},"references\u002Fideation-frames.md",1775,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002Freferences\u002Fideation-frames.md",{"path":91,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":92,"sourceHtmlUrl":93},"references\u002Fissue-tracker-mode.md",1051,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002Freferences\u002Fissue-tracker-mode.md",{"path":95,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":96,"sourceHtmlUrl":97},"references\u002Fpost-ideation-workflow.md",19326,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002Freferences\u002Fpost-ideation-workflow.md",{"path":99,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":100,"sourceHtmlUrl":101},"references\u002Fsubject-identification.md",4051,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002Freferences\u002Fsubject-identification.md",{"path":103,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":104,"sourceHtmlUrl":105},"references\u002Fsurprise-me-mode.md",2479,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002Freferences\u002Fsurprise-me-mode.md",{"path":107,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":108,"sourceHtmlUrl":109},"references\u002Funiversal-ideation.md",12787,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002Freferences\u002Funiversal-ideation.md",{"path":111,"resourceType":74,"resourceTypeDesc":75,"sizeBytes":112,"sourceHtmlUrl":113},"references\u002Fweb-research-cache.md",4130,"https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-ideate\u002Freferences\u002Fweb-research-cache.md",[115,128,142,154,163,172],{"skillNo":116,"name":117,"nameZh":6,"description":118,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":119,"categories":120,"keywords":124,"publishedAt":43},"01M71QQY0KE1184AF58EC1J5C1","skill-one","Sample skill","https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fsample-plugin\u002Fskills\u002Fskill-one\u002FSKILL.md",[121],{"categoryNo":122,"categoryName":123,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发",[29,125,126,127],"skill one","sample","sample skill",{"skillNo":129,"name":130,"nameZh":6,"description":131,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":132,"categories":133,"keywords":135,"publishedAt":43},"01M71QQW3PNK9AYB5HBMAPM72J","disabled-skill","A skill with model invocation disabled","https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fsample-plugin\u002Fskills\u002Fdisabled-skill\u002FSKILL.md",[134],{"categoryNo":122,"categoryName":123,"taxonomyType":21,"taxonomyTypeDesc":22},[136,137,138,139,140,141],"disabled skill","a skill","with model","with","true","model",{"skillNo":143,"name":144,"nameZh":6,"description":145,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":146,"categories":147,"keywords":149,"publishedAt":43},"01M71QQSY02X7DNZQRAGJ6YFGY","claude-only-skill","A skill restricted to Claude Code only","https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fsample-plugin\u002Fskills\u002Fclaude-only-skill\u002FSKILL.md",[148],{"categoryNo":122,"categoryName":123,"taxonomyType":21,"taxonomyTypeDesc":22},[150,151,152,137,153],"claude only","claude only skill","only","code",{"skillNo":155,"name":156,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":157,"categories":158,"keywords":160,"publishedAt":43},"01M71QQRPZ76X623V5P7T1E3DD","default-skill","https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fcustom-paths\u002Fskills\u002Fdefault-skill\u002FSKILL.md",[159],{"categoryNo":122,"categoryName":123,"taxonomyType":21,"taxonomyTypeDesc":22},[161,162],"default","default skill",{"skillNo":164,"name":165,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":166,"categories":167,"keywords":169,"publishedAt":43},"01M71QQQG3FF2HFQHVH8VHA08R","custom-skill","https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fcustom-paths\u002Fcustom-skills\u002Fcustom-skill\u002FSKILL.md",[168],{"categoryNo":122,"categoryName":123,"taxonomyType":21,"taxonomyTypeDesc":22},[170,171],"custom skill","custom",{"skillNo":173,"name":174,"nameZh":6,"description":175,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":176,"categories":177,"keywords":178,"publishedAt":43},"01M71QQMBNC76141GHXBSMENSH","method-write","Drafts, revises, and audits evidence-bound manuscripts, sections, abstracts, and reviewer responses with claim provenance. Use when scientific prose must be written or revised (entry point).","https:\u002F\u002Fgithub.com\u002Fsajor2000\u002Fmethodsmith\u002Fblob\u002Fmain\u002Fplugins\u002Fmethodsmith\u002Fskills\u002Fmethod-write\u002FSKILL.md",[],[179,180,181,182,183,184,30,41,185,186,139,187,188,189],"write","method write","point","evidence","entry","claim","provenance","reviewer","scientific","prose","drafts"]