A skill your agent uses when the final answer strategy calls for a long-form report, publication-quality research writeup, comprehensive analysis, deep dive, whitepaper-style narrative, or detailed…

Apache-2.0Auto-check passedDocuments & Office

Install Long Form Report Writer

skills CLI
$ npx skills add NVIDIA-AI-Blueprints/deep-researcher-agent --skill long-form-report-writer -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install NVIDIA-AI-Blueprints/deep-researcher-agent long-form-report-writer --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/NVIDIA-AI-Blueprints/deep-researcher-agent.git skills-src && mkdir -p .claude/skills && cp -r skills-src/src/deep_researcher_agent/agents/deep_researcher/skills/synthesis/long-form-report-writer .claude/skills/long-form-report-writer && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
long-form-report-writer
GitHub stars
886
Token cost
~2.4k tokens
SKILL.md length
1,194 words
Files
1
Skills in repo
15
Repo updated
First seen
Licence
Apache-2.0

At a glance

A skill your agent uses when the final answer strategy calls for a long-form report, publication-quality research writeup, comprehensive analysis, deep dive, whitepaper-style narrative, or detailed…

  • Works in 9 steps: Confirm the report follows the plan… → Confirm every required answer component… → Confirm every material factual claim has… → …
  • The final answer strategy calls for a long-form report
  • SKILL.md covers When To Use, Long-Form Planning, Report Writing and Report Presentation, plus 3 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Long Form Report Writer is an agent skill from NVIDIA-AI-Blueprints/deep-researcher-agent. Use this skill when the final answer strategy calls for a long-form report, publication-quality research writeup, comprehensive analysis, deep dive, whitepaper-style narrative, or detailed cited Markdown report. Triggers: "longformreport", "long-form report", "comprehensive report", "publication-ready", "deep dive", "in-depth analysis", "whitepaper", "research report", "detailed report", "full report". Outputs: A polished, cited Markdown report with coherent structure, analytical depth, source-grounded claims…

Its SKILL.md is about 2.4k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Documents & Office, covering Report writing, Data visualization and Deep research. The repository describes itself as: The NVIDIA Deep Researcher Agent Blueprint is an open reference example for building intelligent AI agents that connect to your enterprise data, reason using state-of-the-art… The licence is Apache-2.0.

When your agent uses it

  • The final answer strategy calls for a long-form report
  • Publication-quality research writeup
  • Comprehensive analysis
  • Whitepaper-style narrative

Example prompts

  • “longformreport”
  • “long-form report”
  • “comprehensive report”
  • “/long-form-report-writer”

Workflow steps

9 steps, taken from the first numbered list in SKILL.md.

  1. Confirm the report follows the plan constraints.
  2. Confirm every required answer component is covered or explicitly named as a gap.
  3. Confirm every material factual claim has an inline citation.
  4. Confirm the Sources section includes every cited source and no uncited filler. Each source line must contain one full verified URL or one…
  5. Confirm the report does not mention internal files, agents, prompts, or tools.
  6. Confirm internal evidence_judgment scores and rationales are not exposed unless the user explicitly asked for methodology.
  7. Write the complete Markdown answer to /shared/output.md.
  8. Return only the short completion marker Wrote /shared/output.md; do not echo the full Markdown.
  9. Do not use edit loops or search-and-replace passes to repair citation numbering. Deterministic post-processing verifies and sanitizes…

What it can do on your machine

Read from SKILL.md and the folder at commit 951a1a1. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are markdown).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Long Form Report Writer loads about 2.4k tokens when it runs. Until then it costs about 144 tokens; SKILL.md has 1,194 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~144
When it runs · the whole SKILL.md, loaded when a task matches
~2.4k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from NVIDIA-AI-Blueprints/deep-researcher-agent at commit 951a1a1, republished under its Apache-2.0 licence (© NVIDIA-AI-Blueprints). 1,194 words, ~2,360 tokens.

Download SKILL.mdSave it as .claude/skills/long-form-report-writer/SKILL.md (or your agent's skills folder).
name
long-form-report-writer
description
Use this skill when the final answer strategy calls for a long-form report, publication-quality research writeup, comprehensive analysis, deep dive, whitepaper-style narrative, or detailed cited Markdown report. Triggers: "long_form_report", "long-form report", "comprehensive report", "publication-ready", "deep dive", "in-depth analysis", "whitepaper", "research report", "detailed report", "full report". Outputs: A polished, cited Markdown report with coherent structure, analytical depth, source-grounded claims, limitations, and a sources section.

Long-Form Report Writer Skill

Write a publication-quality, well-cited Markdown report from the planned answer strategy and ResearchNotes artifacts. This skill is for final synthesis only; do not perform new research.

When To Use

Use this skill when answer_strategy.answer_type is long_form_report, or when the user's request clearly asks for a comprehensive report, deep analysis, publication-ready writeup, or whitepaper-style answer.

Do not use this skill for brief answers, primary table outputs, data extraction, multiple-choice selection, or numeric predictions unless the answer strategy also requires a full report narrative.

If answer_strategy.answer_type is prediction, or the user asks for a forecast, stock price, price target, probability, odds, expected value, or next-period value, this skill is not the controlling skill. If prediction-report-writer is available, read and follow that skill instead.

Long-Form Planning

Before writing, complete the general cross-synthesis pass from the base writer prompt. Then use think to turn that synthesis map into a report plan:

  • Choose the report title, main sections, and subsection flow from the answer_strategy, the user's request, and the evidence.
  • Use each note's ResearchNotes.evidence_judgment when present to prioritize notes: high-score/high-confidence notes should anchor the report, medium notes can support or nuance it, and low-score or low-confidence notes should mainly inform gaps, caveats, conflicts, or weak-evidence warnings.
  • Decide which components need detailed prose, which need compact summary, and which need explicit caveat or gap treatment.
  • Treat required_components as a coverage checklist, not a mechanical table of contents. Combine, split, or reorder components when that produces a more coherent reader-facing argument.
  • Decide where report-specific elements such as tables, equations, timelines, case-study boxes, or comparison matrices materially improve the reader's understanding.
  • Decide how to order the report so it moves from fundamentals to implications without exposing internal workflow steps.

Report Writing

Write a comprehensive, in-depth report following the answer strategy and satisfying the constraints. If you need to recall research findings during writing, re-read the relevant research-note JSON files before drafting.

For broad explanatory reports, target 3000-5000+ words unless the user or plan asks for a shorter answer. Each section should have detailed paragraphs, not just a few sentences. Ensure all aspects of the user's query are addressed. Provide analytical depth: explain mechanisms and causes, not just surface descriptions. Acknowledge trade-offs, limitations, uncertainty, or open questions where the evidence warrants.

Create a coherent narrative that integrates information across sources and sections. Identify core concepts that appear across multiple notes, recognize complementary findings that build a fuller picture, prioritize recent and high-quality evidence, and connect concrete details to the report's larger argument. Do not produce a sequence of short, isolated bullet points or one-fact paragraphs.

Report Presentation

  • Use clear Markdown headings: # for the title, ## for main sections, and ### for subsections when useful.
  • Write in developed paragraphs for readability; avoid excessive bullet points.
  • Use bullets only for genuinely list-like material such as taxonomies, compact takeaways, risks, or checklists, and surround them with prose that explains their significance.
  • No self-referential language such as "I found" or "I researched".
  • Each paragraph should be substantive and detailed.
  • Structure from fundamentals to complex concepts.
  • Provide contextual explanation of specialized terms.
  • When appropriate, prefer clear language over jargon; explain technical terms when used.
  • Use tables, equations, or code blocks when they materially improve the report.
  • Keep the report human-facing; do not expose the internal planning workflow.

Content To Never Include

Never include:

  • References to agents, constraints, workflow, prompts, tools, or internal files.
  • Methodology sections or meta-commentary about how the report was produced.
  • Statements like "the user requested" or "this report satisfies".
  • Word counts or length statements.
  • Internal workflow tool names such as get_verified_sources, write_file, read_file, think, task, or run_research_batch.
  • Descriptions of how sources were gathered, verified, or validated.

Exception: in the Sources section, a verified data-source tool name may appear when the source has no URL or document locator. Preserve the raw tool/source name exactly as shown by verified sources.

The report must read as if written by a professional human researcher.

Show full SKILL.md (543 more words)Show less

Citation Guidelines

  • Number sources sequentially with inline citations like [1], [2], etc.
  • Place citations immediately following the relevant claim.
  • Citations are required for every material source-derived factual claim, number, date, sourced comparison, source-specific caveat, or analytical conclusion grounded in evidence.
  • For arithmetic, rankings, ranges, means, medians, or summary statistics computed solely from already cited values in the report, cite the underlying source data where those values appear. Then state that the derived statistics are computed from that cited table or cited values; do not attach source numbers to every computed-statistic line.
  • In a summary-statistics block, cite only source-derived inputs, source-specific caveats, or any newly introduced external value. Lines such as count, mean, median, range, min, or max do not need their own citation when they are computed from the immediately preceding cited table or cited value set.
  • Never place bare URLs or hyperlinks in the report body; use only [N] citations inline. URLs belong exclusively in the Sources section.
  • For information supported by multiple sources, use adjacent citations like [1][4][7].
  • Include the full source list at the end of the document.
  • Build the final citation map from the default compact get_verified_sources output. ResearchNotes can guide what evidence to use, but ResearchNotes titles, archive names, publisher names, source labels, and note-local citation numbers are not valid final citations unless the exact same URL or citation key appears in get_verified_sources.
  • Call get_verified_sources(mode="full") only if a source locator from ResearchNotes is missing from the compact output and is materially needed.
  • Assign each unique verified URL or verified citation key a single citation number across all findings. Each citation number must map to exactly one verified URL or one exact verified citation key from get_verified_sources.
  • Do not group multiple publishers, tools, URLs, archive names, collections, documents, or source names under one citation number. Do not write unverifiable aggregate labels
  • Number sources sequentially without gaps.
  • For internal documents, cite the filename/page locator only when the exact locator appears in get_verified_sources.
  • For URL-less structured tool sources, cite the raw tool/source name exactly as shown by get_verified_sources; do not invent friendly titles.
  • Do not invent, recall, paraphrase, shorten, summarize, or "prettify" URLs or citation keys. Use only source locators copied from get_verified_sources.
  • If a claim cannot be mapped to a get_verified_sources entry, remove it or state it as an evidence gap without pretending it is cited.
Sources Section Format

Use this format:

markdown
## Sources
[1] Source Title: https://example.com/source
[2] internal-report.pdf, p.15
[3] mcp_time__get_current_time

Final Verification

Before finishing:

  1. Confirm the report follows the plan constraints.
  2. Confirm every required answer component is covered or explicitly named as a gap.
  3. Confirm every material factual claim has an inline citation.
  4. Confirm the Sources section includes every cited source and no uncited filler. Each source line must contain one full verified URL or one exact citation key copied from get_verified_sources; descriptive source labels alone are invalid.
  5. Confirm the report does not mention internal files, agents, prompts, or tools.
  6. Confirm internal evidence_judgment scores and rationales are not exposed unless the user explicitly asked for methodology.
  7. Write the complete Markdown answer to /shared/output.md.
  8. Return only the short completion marker Wrote /shared/output.md; do not echo the full Markdown.
  9. Do not use edit loops or search-and-replace passes to repair citation numbering. Deterministic post-processing verifies and sanitizes citations after /shared/output.md is written.

© NVIDIA-AI-Blueprints, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in src/deep_researcher_agent/agents/deep_researcher/skills/synthesis/long-form-report-writer of NVIDIA-AI-Blueprints/deep-researcher-agent.

Open the folder on GitHubat commit 951a1a1

Compare with similar skills

Long Form Report Writer next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Long Form Report Writer compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Long Form Report Writer this skillNVIDIA-AI-Blueprints/deep-researcher-agent886—~2.4kAutomated safety check: PassApache-2.0
Long Form Content Frameworksrampstackco/claude-skills945—~7.5kAutomated safety check: PassMIT
Live Researchbrightdata/skills264—~1.8kAutomated safety check: PassMIT
CSV To Executive Reportskrun-dev/skrun210—~1.1kAutomated safety check: PassMIT
Policy Keyword Trackerzhaohui-yang/official-document-drafting140—~555Automated safety check: PassCustom licence
Grounded Summarygaotiexinqu/OneResearchClaw450—~5.1kAutomated safety check: PassMIT

Similar skills

  • Long Form Content Frameworks

    rampstackco/claude-skills

    Patterns for individual long-form content pieces. An agent skill from rampstackco/claude-skills.

    945 GitHub stars~7.5k tokensUpdated 4 days ago
    Documents & OfficeAuto-check passed
  • Live Research

    brightdata/skills

    Produce a deep, multi-source, cited research brief on a topic from live web data using Bright Data's Discover API (intent-ranked web search + parsed page content).

    264 GitHub stars~1.8k tokensUpdated 4 days ago
    Research & ScienceAuto-check passed
  • Turn a CSV of operational data (sales, usage, signups, support tickets) into a multi-page styled PDF executive report with narrative + matplotlib charts.

    210 GitHub stars~1.1k tokensUpdated 19 days ago
    Documents & OfficeAuto-check passed
  • Policy Keyword Tracker

    zhaohui-yang/official-document-drafting

    Collects policies and developments on one topic from Chinese central ministries and compiles them into a verified situation report, with optional Word export.

    140 GitHub stars~555 tokensUpdated 1 mo ago
    Documents & OfficeAuto-check passed
  • Grounded Summary

    gaotiexinqu/OneResearchClaw

    Create a rich, evidence-preserving research report draft from a grounded note and its follow-up literature result.

    450 GitHub stars~5.1k tokensUpdated 5 mo ago
    Documents & OfficeAuto-check passed
  • Research

    iusztinpaul/ai-research-os-workshop

    Build, extend, AND query a persistent LLM-maintained wiki for any research topic.

    179 GitHub stars~17k tokensUpdated 3 mo ago
    Knowledge ManagementAuto-check passed

More from NVIDIA-AI-Blueprints/deep-researcher-agent

All 15 skills in this repo
  • Deep Researcher Configure Workflow

    NVIDIA-AI-Blueprints/deep-researcher-agent

    A skill your agent uses when composing, adapting, or validating an Deep Researcher Agent workflow YAML under configs/ — selecting a shipped profile, enabling tools and datasourceregistry sources…

    886 GitHub stars~1.1k tokensUpdated 2 days ago
    Auto-check: notes
  • Deep Researcher Research

    NVIDIA-AI-Blueprints/deep-researcher-agent

    A skill your agent uses when asked to run deep research or Deep Researcher Agent research through a reachable NVIDIA Deep Researcher Agent Blueprint backend.

    886 GitHub stars~4.4k tokensUpdated 2 days ago
    Auto-check: notes
  • Deep Researcher Add Data Source

    NVIDIA-AI-Blueprints/deep-researcher-agent

    A skill your agent uses when adding or changing an Deep Researcher Agent data source under sources/, registering it as a NeMo Agent Toolkit function, wiring it into the datasourceregistry for UI…

    886 GitHub stars~1.1k tokensUpdated 2 days ago
    Auto-check: notes
  • Deep Researcher Add Tool

    NVIDIA-AI-Blueprints/deep-researcher-agent

    A skill your agent uses when adding or changing a general-purpose Deep Researcher Agent tool (a NeMo Agent Toolkit function) under sources/, defining its FunctionBaseConfig schema, registering it…

    886 GitHub stars~1.1k tokensUpdated 2 days ago
    Auto-check: notes
  • Deep Researcher Customize Prompts Models

    NVIDIA-AI-Blueprints/deep-researcher-agent

    A skill your agent uses when customizing Deep Researcher Agent behavior through Jinja2 prompt templates or per-agent model selection — editing prompts under src/deepresearcheragent/agents//prompts/…

    886 GitHub stars~1.6k tokensUpdated 2 days ago
    Auto-check: notes
  • Deep Researcher Deploy

    NVIDIA-AI-Blueprints/deep-researcher-agent

    A skill your agent uses when asked to install, deploy, run, validate, troubleshoot, or stop NVIDIA Deep Researcher Agent Blueprint infrastructure.

    886 GitHub stars~3.5k tokensUpdated 2 days ago
    Auto-check: notes

Questions about Long Form Report Writer

What does Long Form Report Writer do?

A skill your agent uses when the final answer strategy calls for a long-form report, publication-quality research writeup, comprehensive analysis, deep dive, whitepaper-style narrative, or detailed…. Long Form Report Writer is an agent skill from NVIDIA-AI-Blueprints/deep-researcher-agent. Use this skill when the final answer strategy calls for a long-form report, publication-quality research writeup, comprehensive analysis, deep dive, whitepaper-style narrative, or detailed cited Markdown report.

When should I use Long Form Report Writer?

Long Form Report Writer fits situations like: the final answer strategy calls for a long-form report; publication-quality research writeup; comprehensive analysis; whitepaper-style narrative.

How do I install Long Form Report Writer in Claude Code?

Run `npx skills add NVIDIA-AI-Blueprints/deep-researcher-agent --skill long-form-report-writer -a claude-code`. Or copy the skill folder (src/deep_researcher_agent/agents/deep_researcher/skills/synthesis/long-form-report-writer in NVIDIA-AI-Blueprints/deep-researcher-agent) into .claude/skills/long-form-report-writer in your project. Claude Code loads it when a task matches its description.

How do I install Long Form Report Writer in Codex?

Run `npx skills add NVIDIA-AI-Blueprints/deep-researcher-agent --skill long-form-report-writer -a codex`. Or copy the skill folder (src/deep_researcher_agent/agents/deep_researcher/skills/synthesis/long-form-report-writer in NVIDIA-AI-Blueprints/deep-researcher-agent) into .agents/skills/long-form-report-writer in your project. Codex loads it when a task matches its description.

Can I use Long Form Report Writer in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add NVIDIA-AI-Blueprints/deep-researcher-agent --skill long-form-report-writer -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/long-form-report-writer, .gemini/skills/long-form-report-writer, .github/skills/long-form-report-writer and .opencode/skills/long-form-report-writer in your project.

What does Long Form Report Writer need to run?

SKILL.md names no scripts, command-line tools or credentials: Long Form Report Writer is instructions for the agent only.

Does Long Form Report Writer access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Long Form Report Writer safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Long Form Report Writer use?

Long Form Report Writer is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Long Form Report Writer use?

About 2.4k tokens (SKILL.md is roughly 9.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Long Form Report Writer?

Skills that share tags, products or a category with Long Form Report Writer: Long Form Content Frameworks (rampstackco/claude-skills, 945 stars), Live Research (brightdata/skills, 264 stars), CSV To Executive Report (skrun-dev/skrun, 210 stars) and Policy Keyword Tracker (zhaohui-yang/official-document-drafting, 140 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Long Form Report Writer?

NVIDIA-AI-Blueprints (a GitHub organization) maintains it in NVIDIA-AI-Blueprints/deep-researcher-agent, which has 886 GitHub stars. The repository holds 15 skills in this directory. The repository was last updated on October 9, 2026.

Source: NVIDIA-AI-Blueprints/deep-researcher-agent on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.