Investment Research Agent

Investment Research Agent

Use the active host—Codex or Claude/Cowork (including Claude Code plugin runtimes)—as the reasoning runtime and the bundled research_search, research_get_chunk, and research_get_report_context MCP tools as the corpus interface. The host's existing account supplies the model. On the first research call, YouWare opens a browser for account sign-in and device authorization when needed. A new workspace's first retrieval returns a quote without charging; explicit confirmation starts its run and consumes one research credit, or zero credits during active annual unlimited access. All later retrieval using that same absolute workspace while the run is active is included. In all user-facing updates, refer to the product and research capability only as YouWare; never expose legacy product names or internal compatibility identifiers.

Do not write a long report in one pass. Keep evidence.json as the unified evidence backbone. MCP is the shared data plane only: it must never spawn an agent, invoke a model CLI, or call a model API.

Concise Start Gate

This is a long-running workflow that normally requires at least one hour. Before creating the workspace, reading task-specific source material, calling any research or retrieval tool, starting a remote run, or dispatching research workers, send one concise paragraph in the user's language. Lead with the decision value the work will create—an independently tested investment conclusion, traceable multi-source evidence, an explicit market-expectations gap, reproducible valuation, and actionable monitoring or invalidation conditions—rather than leading with elapsed time. Then give a realistic duration range whose lower bound is at least 60 minutes, the high-level research plan, the use of professional sell-side brokerage research and industry-expert interview databases, and the two final deliverables. State that unusually complex, poorly disclosed, or multi-company topics may take longer. End with a short, natural choice: invite the user to reply with a clear start confirmation when the scope and timing work, or to state any requested changes to research focus, investment horizon, or deliverable format. Say naturally that research begins after their confirmation; do not end with a vague yes/no question. Then wait for explicit user confirmation. This value-and-timing notice is a hard preflight gate, not an optional progress update: do not begin execution merely because the user supplied a research topic, and do not defer the notice until after tools have already run.

Use this Chinese default when appropriate, adapting the company, scope, and time range without making it longer:

我会交付一份可直接支持投资决策的机构级深度研究：用券商研报、产业专家访谈与官方披露交叉验证核心假设，拆出市场预期与预期差，建立可复算的经营预测和情景估值，并给出投资行动、风险传导与证伪条件。这项工作通常需要 60—120 分钟，复杂标的可能更久；最终交付 HTML 和 Markdown 两份专业报告。如果以上安排符合你的预期，请回复“开始”；如果希望调整研究重点、投资期限或交付方式，也可以直接告诉我。我会在你确认后开始研究。

Treat “确认”“开始”“按这个计划做” and clear equivalents in the same conversation as approval. If the user's initiating message already explicitly says to start immediately without another confirmation, the start gate is satisfied, but still send the concise timing-and-plan paragraph before the first tool call. Do not show internal agent roles, staged filenames, tool names, retrieval mechanics, backend phases, run state, persistence state, billing mechanics, or a long checklist. If the Finance service later returns a priced confirmation quote that was not available at the start gate, show only the topic, cost, and balance and obtain the required cost confirmation; do not repeat the research plan. After confirmation succeeds, continue research silently: never announce that a run started, that state was saved, that the user was or was not charged, that access is included in a plan, or that an internal research phase has begun.

Expert Evidence Gate

sell and primary may be searched in either order. Keep sell and primary as separate corpus searches. Primary results have no expert/official source label, so within primary run separate expert-targeted and official-targeted research_search calls using different concepts and phrases—never source_types. Search experts with company or product plus customer, supplier, competitor, channel, and former-employee roles, varying demand, orders, inventory, capacity, pricing, delivery, product progress, share, and time terms. Finding official materials does not complete primary research. Expert evidence is the highest-priority retrieval requirement for completed research: the report must show multiple independent expert passages distributed across the core industry claims rather than concentrated in one section. This is a floor on expert coverage, not a ceiling on other evidence — never cut or thin sell-side or official content to shift the evidence mix toward experts. If a core industry claim lacks expert evidence, repeat the expert search with different roles, chain positions, and wording. If a gap remains after exhaustive expert search, still write the complete report: state the gap explicitly where the affected claim is argued and label that claim as lacking expert corroboration, instead of dropping the claim, thinning the analysis, or withholding the report because only official material supports it.

Cross-Runtime Run

1. Create a project workspace at .bloome/research/<topic-slug>/.
2. Use the bundled research tools for retrieval and the active host for planning, validation, and synthesis. For any listed security, also read the bundled ../stock-market-data/SKILL.md and use its unified dispatcher for security identity, current price, and historical market data.
3. Save every required staged artifact in that workspace.
4. Call validate_research_workspace before final delivery and repair every reported error.
5. Call open_research_workspace with the absolute workspace path. In Codex, promote the compact launcher into the native PiP panel and use fullscreen for the report. In Claude Code, use the returned reportPath to inspect report.html; the same progress, evidence, artifact, and validation data remain available without a rendered MCP App panel. The workbench may frame the report where supported, but the report itself must use the bundled React static renderer, which reads assets/template.html as its visual source of truth.

Useful starter requests:

研究 AI 推理需求对 NAND 价格周期的影响，并生成完整研报。
打开这个项目的 YouWare 投资研究工作台。
验证当前研报是否满足 investment research 的全部输出要求。

Before the first corpus tool call in each task, tell the user in their language: YouWare may open in the browser for sign-in and device approval on first use, and research will continue automatically after approval. Never open an unexplained login window.

Billing Confirmation Gate

confirmationRequired is a quote, not an error, quota exhaustion, or an execution-environment block. When a retrieval tool returns it:

1. Stop and show the returned topic, cost, and current balance.
2. Ask the user to confirm the quoted cost in conversation. Do not substitute stale workspace chunks or claim remote retrieval is unavailable while confirmation is pending.
3. Call confirm_research_run only after the user explicitly agrees.
4. After confirmation succeeds, retry the original retrieval with the exact same absolute workspace path. Do not ask again during that active run.

The confirmation result is operational metadata, not a user-facing milestone. On success, do not summarize or quote it; immediately resume the previously described research. User-visible progress updates may state the substantive question being investigated or a material finding, but must not expose run activation, saved state, billing outcome, credit handling, tool calls, corpus labels, or internal stage names such as landscape search.

Only a 402 response means the account has no usable credits. Report that exact status and direct the user to the returned pricing URL. Pass the same absolute research workspace path in every corpus tool call; changing it creates a separate quoted run. The local MCP stores the authorized device credential under ~/.bloome/ and a non-secret run marker inside the workspace. Once the finished report passes validation, validate_research_workspace generates and returns its directly accessible /reports/<id> link while closing the active run. Keep user-facing delivery simple: before the action, say “我现在生成报告链接。” only when a progress update is useful; after success, say “报告链接已生成：”. Do not narrate transport, storage, or publication infrastructure; describe the result only as a directly accessible report link. The approved research run already authorizes final link generation, so no second confirmation is needed. If material new evidence is found before link generation, update and revalidate the report first; never knowingly generate a stale link and explain the missing evidence as a caveat afterward. If authorization is revoked, credits are exhausted, the gateway is unavailable, or search returns no results, preserve partial artifacts, report the exact retrieval status, and stop evidence-based conclusions. BLOOME_FINANCE_URL may override the Finance service URL for local development.

Parent and Subagent Roles

The parent is the research lead and final editor. It owns the landscape pass, module plan, evidence reconciliation, decision, outline, final editing, visual review, and validation. Evidence workers produce module memos; after evidence is frozen, chapter workers each write one planned chapter. No worker writes a shared file.

After the landscape pass, save plan.json with enough non-overlapping modules to cover the topic deeply, using the fields defined in references/module-contract.md. Let the question determine module count, but when coverage depth is in doubt, prefer more, narrower modules over fewer broad ones: split any module whose question contains two separable evidence questions, and give each mechanism, market layer, or company that materially affects the conclusion its own module. Prefer host-native delegation:

• Claude/Cowork: delegate evidence scopes to the bundled research-module subagent, chapter scopes to the bundled chapter-writer subagent, and optionally use evidence-auditor after all memos exist.
• Codex: use native subagents with the same module, chapter, and auditor contracts. Do not require users to install custom .codex/agents files.

Let the host manage worker scheduling and concurrency. In the evidence pass, each worker handles one scope and writes only modules/<id>.md. In the writing pass, each worker handles one outline section and writes only its assigned chapter_XX_*.md. If native subagents are unavailable, denied, lack the required file or research-tool access, or fail, run only the missing modules or chapters sequentially in the parent with the identical contract. Never replace host delegation with a spawned Claude, Codex, Pi, or model-API process.

Read references/multiagent-workflow.md, references/module-contract.md, and references/chapter-contract.md before planning or dispatching workers.

Required Workflow

Run these stages in order:

1. Search both sell and primary corpora for the initial landscape in either order, passing the same absolute workspace path to every research tool call. Keep each search focused on one evidence question: strings within one concepts group are alternatives, while separate groups narrow the query together. Respect the tool schema (chunks_per_report 1–3 and at most six strings per concept group) instead of learning limits through failed calls. Within primary, make and record separate expert-targeted and official-targeted calls as required above. Handle the one-time conversational research confirmation before continuing retrieval.
2. Save the topic-shaped module plan, dispatch host-native workers or use the sequential fallback, and read every modules/<id>.md memo.
3. Search both corpora iteratively with varied query seeds, relevant time windows, exact chunk reads, and surrounding context. Do not dump several full search responses into one combined tool output; retain compact report/chunk IDs, then fetch only the selected complete passages so host output truncation cannot silently discard evidence. In every primary iteration, keep expert-targeted and official-targeted directions separate; official results do not count as expert coverage. Continue expert retrieval with different roles, value-chain positions, and wording until the core industry claims are covered by multiple independent expert sources. For every retrieval round, write the exact accepted evidence IDs into that round's accepted_evidence_ids in coverage_stats.json; a recorded search with no accepted evidence does not satisfy the evidence gate. A completed institutional report must accept and cite passages from at least three independently authored sell-side documents. Distribute sources by analytical role—such as operating assumptions, industry structure, forecast calibration, disagreement, and valuation—according to what each document actually establishes. Multiple chunks from one underlying document still count as one source. Reusing a strong document is allowed when it is genuinely the best source; repetition must not be presented as independent corroboration or used in place of researching materially different questions. Continue until additional retrieval no longer materially changes the claims, conflicts, or known gaps, or access is exhausted. Record the stopping reason and remaining gaps in coverage_stats.json; do not use record counts as a proxy for depth.
4. Reconcile every module candidate in evidence_disposition.md: accept it into the evidence backbone or reject it with a reason. Then save sell_side_logic.md, validation.md, and the unified evidence.json. Every accepted item must link to one or more claim IDs and state whether it supports, challenges, or contextualizes them.
5. Save decision.md as natural Markdown. For a ranked investment decision, state the priority rule and ranking, compare every alternative on the same basis, explain any normalization, identify the evidence that drives the order, and say what would change it. Use scores or weights only when they improve the reasoning.
6. Save report_outline.md as a natural-language editorial plan using references/report-structure.md. Do not add opaque section or visual IDs. For comparative or ranked decisions, plan enough sections to show the decision rule, common comparison basis, winner's full demand-to-earnings mechanism, cycle/scenario boundary, security-level valuation, every material alternative, monitoring, and unresolved gaps. For each section, make an explicit visual decision: write Planned visual: <descriptive-key> — ... when a visual materially improves the evidence-backed argument, otherwise write Visual treatment: prose — ... with the reason. Choose the component type later in visuals.json; do not duplicate it in the outline. Keys must describe the argument, not use labels such as V01. For full-company institutional research, use the NVIDIA visual benchmark in references/visual-spec.md; usually plan 15–20 useful figures without a hard count gate. Plan sections so that every module whose evidence was accepted is visibly represented in the report; do not average several evidence-rich modules into one thin section.
7. Freeze evidence.json, decision.md, and report_outline.md, then assign one chapter worker to each planned substantive section using references/chapter-contract.md. Give it only the section brief, report language, decision.md, relevant accepted evidence and module memos, and neighboring-section boundaries. Each worker writes one unique chapter_XX_*.md; chapter workers may run concurrently because they never edit shared files. Use natural headings. Every chapter needs a direct answer, source-backed mechanism, relevant numbers or calculations, investment implication, and an explicit boundary, opposing-evidence, risk, or invalidation discussion. Do not let chapter drafts collapse into executive-summary paragraphs. Write the reasoning out at full length: every causal step in a chain gets its own explicit statement with its supporting evidence — never leap from a premise to a conclusion across unstated intermediate links.
8. Edit final_report.md from the chapter drafts in outline order. Start with one explicit H1 report title, followed by consistent H1 or H2 chapter headings. Keep the investment conclusion once as an ordinary prose section with an author-written, conclusion-led heading: no mandatory 核心判断 label, no highlighted judgment card, and no automatic section naming. A compact metadata preamble may follow the title; it is never interpreted as a judgment. Preserve its date, currency, price and horizon. After the judgment, begin the substantive body with company cognition and the business model rather than repeating the recommendation. Use {{cite:<evidence-id>}} for inline evidence so the renderer, not prose generation, owns the displayed title and locator. The parent may rewrite, merge, and de-duplicate chapter prose for flow, but must preserve each chapter heading, direct answer, distinct causal mechanism, decisive number or calculation, exact citation, primary calibration, disagreement, boundary, invalidating condition, ranking-flip condition, and descriptive {{visual:key}} placement marker. Remove repeated background and repeated explanations even when their wording differs. Do not reduce a substantive chapter to an executive-summary paragraph. Before rendering, compare the final report against every chapter using this checklist and restore any missing analytical element.
9. Use the investment-visualization skill and references/visual-spec.md to save every planned visual as a supported, evidence-linked component specification in visuals.json; use { "visuals": [] } only when prose is explicitly clearer. Prefer visualizing every decision-relevant time series, composition, comparison, causal bridge, scenario range, valuation bridge, sensitivity, and monitoring framework for which structure reduces reading effort; this is an expansive visual policy, not permission to duplicate the same fact or decorate unsupported claims. Every quantitative visual must supply a visible unit and direct display labels for every plotted value; every visual must supply an accessible description. Place each visual in final_report.md on its own line as {{visual:<descriptive-key>}}. When two compact visuals answer complementary questions on the same reading step, place their markers consecutively; the renderer lays them out as a two-column row on desktop and one column on narrow screens. Do not pair dense tables, matrices, or visuals that require independent narrative context. Then call render_research_report with the absolute workspace path. The tool atomically synchronizes final_report.md to report.md, then parses Markdown into tokens and renders paragraphs, citations, primary quotations, tables, and visual specifications through controlled React components. Never write page HTML, raw SVG, CSS, or scripts in final_report.md or report.md. The output is self-contained, contains no browser React runtime, and is ready for upload.

Investment reports must not use fenced code blocks, four-space-indented pseudo-tables, ASCII borders, or monospaced audit panels for analytical content. They render as code and overflow in HTML. Convert exact lookup content to a Markdown or controlled table, causal content to flow, and quantitative relationships to the appropriate controlled visual. 10. Inspect the compiled report at desktop and narrow-phone widths, rerun render_research_report after any Markdown or visual change, then call validate_research_workspace. Use open_research_workspace for the native report surface when direct file:// browser access is unavailable; do not replace visual inspection with structural validation or claim screenshots were reviewed when no visual surface was available. Repair every error before delivery. Successful validation also closes the research run.

Keep all staged files traceable to evidence.json. Do not skip from search notes or module memos directly to the final report.

The bundled research_search and research_get_chunk tools are the research corpus interface. Do not infer that the corpus is unavailable merely because no separate “knowledge base” skill is installed. If search returns no results or the research proxy is unavailable, report that exact retrieval status and stop evidence-based conclusions; do not replace the research with unsupported industry generalizations.

Evidence Layers

sell contains research published by sell-side institutions. It represents what the market believes: earnings forecasts, key assumptions, debates, risks, and valuation frameworks. Use it to extract the investment logic, causal chain, key assumptions, forecasts, disagreements, and valuation framework, and to understand what the market has priced in. Treat its conclusions as hypotheses to test against industry-expert and official material, not proof of industry reality.

sell is also the primary source of structured quantitative material—market size, shipments, pricing, capex, shares, costs, forecasts, model tables, and historical series—and it should be used extensively to build the report's tables, charts, forecasts, valuation ranges, and model calculations. Calibrate key figures against primary where available, but using sell quantitative content is expected and correct, not a defect; only its uncalibrated conclusions are treated as hypotheses rather than proven fact.

Current Market Price Gate

Never use a price printed in a sell-side report as the report's current share price, valuation anchor, or basis for upside/downside. It is only historical context for the broker's publication date. For every listed security, use the bundled stock-market-data skill rather than writing an ad hoc yfinance call. First run its unified security resolver (python scripts/market_data.py resolve "<company or ticker>" from that skill directory), then fetch the quote through python scripts/market_data.py price <resolved-ticker>. The dispatcher uses the configured provider chain—yfinance for supported US/HK/JP/KR securities and explicit market-specific fallbacks where configured—and returns provenance and freshness metadata. Do not bypass it with yf.Ticker(), yf.download(), a web snippet, or model memory.

Save the normalized resolver and price outputs in the research workspace as market_data_snapshot.json. The snapshot must contain the resolved entity/listing/security identity, ticker, exchange or venue, currency, price, quote timestamp or latest completed trading date, retrieval timestamp, quote availability/delay status, adjustment convention, provider, source metadata, provider attempts, and whether a fallback was used. Use the current regular-market price when available and clearly timestamp it; otherwise use the latest completed trading-session close. Use the same verified price and valuation timestamp consistently in the opening metadata, reverse valuation, scenario tables, target-price returns, charts, and investment conclusion.

Prefer unadjusted market price for the current valuation anchor and adjusted prices for total-return historical series unless the analysis requires a different documented convention. For ADRs, secondary listings, share classes, or cross-listed securities, verify that the ticker, trading venue, currency, and share/ADR ratio match the security being valued. Respect the dispatcher's declared fallback chain and freshness result. If every configured provider is unavailable or stale, disclose that the current price is unavailable and stop price-based valuation; never fall back to a broker report's quoted price. A listed-security report fails delivery when market_data_snapshot.json is missing, its stated “current price” is supported only by sell-side research, or different sections use inconsistent price dates.

Within primary, search two material categories separately:

• Industry-expert material: expert interviews, former-employee interviews, industry-participant or consultant conversations, channel checks, fieldwork, and research notes based on direct industry-participant commentary. Search across customers and end users, procurement or operations staff, upstream suppliers, competitors, distributors and channel partners, integrators, and former executives or employees. Use these materials to identify leading changes in demand, orders, inventory, capacity, pricing, delivery, product progress, and market share. Seek supporting, opposing, and conflicting views from different roles and value-chain positions.
• Official material: regulatory filings, company announcements, government documents, investor-relations materials, and earnings releases or calls. Use these materials to confirm disclosed facts and management statements. Earnings-call management commentary is official material, not an expert interview.

Finding official materials does not complete primary research. Continue searching industry-expert material until the core industry claims have broad, independent expert coverage or the remaining evidence gap is explicitly reported.

primary outweighs sell. Sell-side views are hypotheses to be tested against primary reality, not co-equal proof: when the two conflict on the same question, let primary control the conclusion and keep the overruled sell-side view visible as a disagreement. A claim supported only by sell-side narrative carries less evidence strength than one calibrated by primary material, and should be labeled accordingly.

For every material claim, record support, opposing evidence, calibration result, unresolved point, evidence strength, and what would change the judgment. If no primary calibration exists, state that plainly.

Long Report

The expert-evidence gate is a quality bar on the evidence, not a scope limit on the report. Passing it does not mean the report is finished: the final report must still present the complete argumentation for every accepted module, covering each accepted module's full reasoning from evidence to conclusion. Satisfying the visible-primary requirements never licenses dropping analysis, chapters, or accepted modules to make validation easier.

final_report.md is an editorial synthesis of complete chapter drafts, not a one-shot answer, an unedited concatenation, or a dump of module memos.

• Let the topic and available evidence determine the final length. Do not set word, character, chapter, argument, or source-count targets. Judge completeness by whether the report contains every decision-relevant layer supported by the evidence.
• Argue completely. Every material conclusion must show its full inference chain — data → mechanism → intermediate inference → conclusion — with each link stated explicitly and either evidence-backed or flagged as an assumption. Do not skip intermediate reasoning steps; if a link cannot be supported, say so instead of writing around it. When brevity and a complete argument conflict, keep the complete argument.
• Before finalizing a deep comparative or ranked report, reopen the outline and module memos and check for missing mechanism, normalization, scenarios, alternatives, valuation, monitoring, or unresolved conflicts. Give each material alternative enough separate treatment to make the ranking auditable; do not hide distinct investment questions inside one compressed paragraph.
• Each substantive section should connect its conclusion to source-backed data, causal transmission, comparison or calculation where relevant, primary calibration, boundary/opposing evidence, and research implication. The winning thesis needs a complete demand → qualified supply → pricing → margin/EPS → valuation bridge.
• Preserve useful detail from chapter drafts: decisive numbers, formula or normalization basis, scenario assumptions, stock-specific catalysts, disconfirming evidence, and ranking-flip conditions. Summarize source descriptions, not the reasoning needed to audit the investment decision.
• Use extracted sell-side data, primary calibration, disagreements, scenarios, sensitivities, and company-level transmission where they help answer the topic. Do not add generic filler or repeat the same number.
• Rewrite and de-duplicate chapter material when it improves flow, including semantically repeated background with different wording. Before delivery, compare final_report.md against every chapter and decision.md; preserve each distinct mechanism, decisive number or calculation, citation, disagreement, boundary, unresolved point, invalidating condition, and the ranking recorded in decision.md. Resolve draft contradictions instead of concatenating them.

Final Report and HTML

Use sell-side material for the analytical frame and structured data; use primary material for validation, narrative proof, and calibration. Sell-side data may be used extensively in tables, charts, forecasts, valuation ranges, and model calculations, but key figures should be calibrated by primary evidence where available. Expert evidence is the report's highest-priority reader-facing evidence layer. The report body must show complete passages from multiple independent expert sources and cover the core industry claims; a single quote, sentence excerpt, or source-only listing is invalid. For every core industry claim supported or challenged by expert evidence, show at least one matched expert passage in a visible primary-quote block.

A completed report must also cite at least one accepted passage from the sell corpus in the body. A bibliography, public-media source list, or an expert-search round without accepted evidence does not count. Validation must trace sell-side and expert retrieval rounds through accepted_evidence_ids to evidence.json, then from evidence.json to the report's inline citation or visible expert quotation.

Write primary evidence quotations as ordinary Markdown blockquotes, followed inside the same blockquote by 来源：<exact source party/title> · <publication date>. The attribution is a compact footer in the same evidence component: never follow a quotation with a separate full-width source-only card. Reader-facing attribution must name the actual institution, expert platform, company, regulator, or publisher plus a recognizable title/date; never expose retrieval keys, chunk IDs, filenames, handles, or strings such as id:official-*, primary_*, sell-*, chunk-*, or source-*. Those identifiers belong only in evidence.json. The React renderer turns valid blockquotes into visible .primary-quote components. Keep page/line locators inside evidence.json for traceability, not in the reader-facing source line. Never write a generic label or invent a source. Write exact sell-side citations in Markdown; the renderer maps them to evidence and creates .src hover tooltips.

Match every reader-facing quotation and source label to the report language. For a Chinese-language report, translate every non-Chinese sell-side tooltip, primary quotation, source title, and descriptive expert role into complete, faithful Chinese before rendering it; do not show a Chinese translation followed by the English original in parentheses or expose a long English passage merely because the source is English. Preserve every sentence, number, date, qualifier, and paragraph boundary, and do not summarize, abridge, or insert ellipses. Store the source passage and original title verbatim in evidence.json.quote and evidence.json.title for auditability; store the full Chinese display translation in evidence.json.quote_zh and the localized reader-facing source title or expert descriptor in evidence.json.title_zh. Render the localized fields in inline citation labels, tooltip headers and bodies, visible primary quotations, visual source lists, and the workbench evidence ledger. When the source passage or title is already Chinese, render the original field directly and omit its _zh counterpart. Institution and platform proper names may remain in their recognized official form, but surrounding source type, report title, expert role, locator, and interface labels must use the report language. Keep tooltip body text plain without <u> or underlines. Every claim, quote, table source, chart source, and tooltip must map to evidence.json. Keep disagreements visible.

Place each visible primary-quote immediately after the paragraph, list item, or table interpretation that states the claim it supports, challenges, or calibrates. Introduce the quotation in the surrounding prose and explain its investment meaning before moving to the next claim. Multiple independent quotations may appear together when they add distinct evidence to the same immediately preceding claim. Do not relocate decision-relevant quotations away from their argument merely to collect them elsewhere in the report. Use only accepted evidence mapped in evidence.json; do not invent or add evidence to satisfy layout or coverage rules.

The bundled React static renderer reads assets/template.html as the visual source of truth and owns the page shell, Markdown token rendering, section wrappers, evidence-linked citation tooltips, responsive rules, controlled visual components, and static-document assembly. Models write report.md plus visuals.json; they must not insert raw page HTML, SVG, CSS, JavaScript, a card layout, or a different structure. Preserve the template's header, ordinary section layout, source bar, hover-tooltip system, and visual language. Do not reintroduce a judgment box, a fixed 核心判断 heading, or a highlighted paragraph between chart groups. Record report_month or the data cutoff in coverage_stats.json so the renderer can display it.

Keep report.html reader-facing and single-page, exactly following the bundled template's structure and visual language. Do not add report/evidence tabs or embed the audit ledger into the page. The audit trail remains in sell_side_logic.md, validation.md, evidence_disposition.md, decision.md, and evidence.json. Do not leave template placeholders unresolved. Treat duplicated opening/core-judgment components, unresolved judgment placeholders, disconnected source-only cards, and reader-visible internal evidence IDs as hard delivery failures.

User-facing Deliverables and Names

Expose exactly two final files to the user: one self-contained HTML report and one Markdown report with identical content. Do not present report.md, final_report.md, evidence ledgers, visual assets, ZIP archives, JSON files, chapter drafts, or other workspace artifacts as additional deliverables. They remain internal research artifacts.

Use a professional, descriptive basename rather than report or final_report:

• Chinese: {公司简称}_{Ticker}_机构级深度投资研究_{YYYY-MM-DD}
• English: {Company}_{Ticker}_Institutional_Investment_Research_{YYYY-MM-DD}
• Japanese: {Company}_{Ticker}_機関投資家向け投資調査_{YYYY-MM-DD}
• Korean: {Company}_{Ticker}_기관투자자용_심층투자리서치_{YYYY-MM-DD}

Use the research cutoff date, preserve a widely recognized company name in the report language, keep the ticker uppercase, replace unsafe filename characters with underscores, and use the same basename for .html and .md. For example: 英伟达_NVDA_机构级深度投资研究_2026-09-01.html and 英伟达_NVDA_机构级深度投资研究_2026-09-01.md. Validation returns professional deliverables paths; use those exact files in the final response. Never attach or link the workspace's internal report.html, even with a renamed link label. In the final response, link only these two files plus the directly accessible report URL when one was generated.

Long HTML should feel like an editorial report, not a tall text dump. Keep the full prose, but break it with natural section labels, compact evidence tables, visible primary quotations, and only the figures that materially advance the decision. On desktop and narrow-phone screenshots, inspect the beginning, middle, and end of the page; verify that wide tables remain readable, no section is clipped, and the renderer has not silently dropped late-report content.

For visual selection, financial chart grammar, annotation, uncertainty, responsive composition, and screenshot-based review, use the separate investment-visualization skill. Keep that editorial judgment in the skill rather than encoding it as validator regex or fixed HTML classes.

Output References

Read references/file-specs.md for staged artifact shapes, evidence fields, and coverage statistics.

Read references/multiagent-workflow.md and references/module-contract.md for host-native delegation, fallback, plan fields, and module memo output.

Read references/report-structure.md for the natural-language outline and chapter order.

Read references/visual-spec.md before writing visuals.json or placing visual markers in report.md.