research-report
Skill1 starsChanged 5 months agoMITView on GitHub
Deep research on a new technology paradigm or trend — what it is, how it is built, its philosophy — plus the providers that sell it and the places it fits in your own codebase, delivered as a clean, visual, team-shareable PDF report. Runs parallel researchers (via a deep-research skill when one is installed), a codebase pass, a report writer, then renders with the pdf-report skill. Triggers on "do a deep research into X and make me a PDF", "research this paradigm, the providers and where we can use it", "what is this new trend and should we adopt it", "write the team a report on X".
14 other projects use a version of this.
Adopted by
Most recently taken up by
- Axect/skillstheir own version, first seen Nov 16, 2026
- EnkrateiaLucca/agent-skills-hands-ontheir own version, first seen Nov 15, 2026
- methylo/claude-skill-booktheir own version, first seen Nov 8, 2026
- Wattson-Law/Xiaobai-Reasearchtheir own version, first seen Nov 8, 2026
- 200iqlabs/shared-skillstheir own version, first seen Oct 31, 2026
- tony/skillstheir own version, first seen Oct 31, 2026
Other versions14 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 14
- PaddlePaddle/FastDeploy★ 3.7ksame name
- Weizhena/Deep-Research-skills★ 2.2ksame name
- ShenSeanChen/waku-agent★ 2.2ksame name
- methylo/claude-skill-book★ 12same name
- guishiw/Guishi-Skills★ 8same name
- 200iqlabs/shared-skills★ 4same name
- Axect/skills★ 4same name
- EnkrateiaLucca/agent-skills-hands-on★ 4same name
- mayuemarsha-del/pm-skills★ 4same name
- Wattson-Law/Xiaobai-Reasearch★ 4same name
4 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository51 other files it gives its agents.
- architect-enterprise-aiSkill
- author-agent-skillsSkill
- build-agent-harnessesSkill
- build-aiewf-2024-ragSkill
- build-generative-mediaSkill
- build-multimodal-productsSkill
- build-rag-search-stacksSkill
- build-robotics-aiSkill
- build-with-deepmindSkill
- build-with-microsoft-aiSkill
- build-with-openaiSkill
- deploy-public-sector-aiSkill
- design-agent-skillsSkill
- evaluate-with-braintrustSkill
- lead-ai-engineering-teamsSkill
- lead-ai-transformationSkill
- learn-summit-2023-remoteSkill
- learn-summit-2023-talksSkill
- navigate-aie-europe-2026Skill
- navigate-aiewf-2024-keynotesSkill
- observe-ai-productionSkill
- operate-agent-memorySkill
- operate-legal-ai-agentsSkill
- operate-openclaw-agentsSkill
- optimize-llm-inferenceSkill
- plan-sovereign-aiSkill
- price-ai-productsSkill
- run-aiewf-2024-evalsSkill
- run-llm-evalsSkill
- run-summit-2023-workshopsSkill
- secure-ai-agentsSkill
- ship-ai-as-pmSkill
- ship-aiewf-2024-agentsSkill
- train-llms-locallySkill
- architecture-reviewSkill
- chat-failure-auditSkill
- create-a-skillSkill
- cross-model-reviewSkill
- cto-weekly-reviewSkill
- friction-auditSkill
- ig-carouselSkill
- pdf-reportSkill
- raise-prSkill
- reel-editSkill
- ship-checkSkill
- spec-grounded-designSkill
- state-lifetime-decisionSkill
- ste100Skill
- scroll-engineeringSkill
- ui-uxSkill
- verify-like-a-userSkill
The file
We split this skill into the rules it gives an agent
1 rule. 0 are common, 0 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 11 edited versions we read keep these.
0
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 43
Research in parallel: If a deep-research skill is installed (e.g.
“Common” is measured: kept by at least 6 in 10 of the 11 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Research Report — paradigm → providers → your opportunities → visual PDF
Arguments: the topic, plus any products, companies or providers the user named
(e.g. always-on sandboxes; products: Grok Bot, ChatGPT dots; providers: DigitalOcean, Daytona).
Output: {SKILL_OUTPUT_DIR}/research-report/{slug}/ — notes/, {slug}.md, {slug}.html, {slug}.pdf.
The brief
There is a new paradigm we are seeing: {topic}, introduced or popularised by {named products/companies}. Research it very deeply: how these systems are built, how they work, and what their philosophy is.
Cover which providers actually offer it to companies and startups like us (start from {providers}) and which are best for us.
Then read our codebase, understand the product, and name the places and opportunities where we should use it to get ahead.
Finally make a very clean PDF with visuals: every number or statistic that can be visualised is visualised; what reads best as text stays text. Sections like a report the team can read and share.
Never add duplication. Nothing appears twice — no heading, section title, chart, table, caption, callout or point is repeated if it is already in the report. Before rendering, scan for repeated headings, the same caveat sentence above and below a chart, and tables that restate a chart without adding columns; keep one.
Process
Research in parallel. If a deep-research skill is installed (e.g. Anthropic's
deep-research), invoke it and act as its coordinator; otherwise spawn subagents directly, or run the five passes in sequence. Five researchers:- Products & philosophy — verify exact names, companies, launch dates, prices. Users often dictate names wrong; find the real ones and record them for a "name check" callout. Include precursors and peers, dated.
- Architecture — how it is built, with numbers (latencies, limits, costs), the security model, published reference architectures.
- Providers A and Providers B — split the list. Per provider: offer, limits, pricing with a worked monthly cost for one reference unit, region/residency and compliance, SDK fit for your stack, maturity. A comparison table ready for charts.
- Your codebase — read-only: what exists today (file paths), pain points the paradigm solves, ranked opportunities with value / effort / risk, exact numbers from code. Never modify the repo, stash, or touch production.
Give every researcher today's date, the current-state constraint, and the instruction to write its notes file incrementally so an interrupted session loses nothing.
- Write the report. Executive summary; then three sections — the paradigm, architecture and providers (ending in a recommendation for you), opportunities in your product (with risks and a phased roadmap); appendix of caveats and sources. Every chartable number goes in a clean markdown table. Claims sourced; estimates and vendor claims labelled. Plain language.
Render the PDF with the
pdf-reportskill — it owns the HTML → headless Chrome pipeline, print CSS and inline SVG. Every chart uses numbers from the report's tables, never invented ones. Charts worth having for this kind of report:- launch timeline of the products and peers
- cost per provider (grouped bars: always-running vs typical usage)
- limits / timeouts per provider; latency on a log scale, vendor claim vs your measurement
- provider fit matrix with rating chips
- an architecture diagram and a lifecycle/state diagram
- opportunity value-vs-effort scatter; a gantt roadmap
Axis units and a source / "estimate" footnote under every chart. A cover, a contents page with page numbers, and an executive summary with 3–5 stat callouts.
- Verify by looking, then fix. See verify.md for the page-by-page check and the traps specific to long, chart-heavy reports.
- Deliver the PDF with a short summary: page count, what each section shows, and which numbers are estimates or unverified. Never delete user files; temp files go to a scratch directory.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

