Automated summary
Research assistant that outlines an open-ended question, triangulates at least three sources, and writes a brief with an explicit confidence rating.
Strengths
- Forces a user-approved outline before any deep searching begins
- Requires at least three independent sources and names both sides when they conflict
- Labels every claim as [primary], [secondary] or [inferred] so deductions are never passed off as fact
- Closes with a justified Confidence section and admits paywall or proprietary-data blind spots
Limitations
- Entirely dependent on the quality of the web_search_tool and http_request tools available in the host environment
- No prescribed citation format (URL, date, author), which can hurt verifiability
- Storage only happens on explicit request, so briefs are not retained by default
Use it when an open-ended question needs a multi-source written brief rather than a one-shot answer.
Skip it for simple factual lookups or in environments where no web access is available.
Security analysis
CautionThe content is purely instructional — no scripts, encoded payloads, destructive commands, credential handling, or attempts to override the agent's role. The only meaningful concerns are the undocumented tool surface (network fetch + persistent memory) and the mismatch between the listed tools and the empty declared_allowed_tools field. Network use is legitimate for research, so this lands at caution rather than risky.
- •Tool declaration mismatch: the skill body lists web_search_tool, http_request, and memory_store, but declared_allowed_tools is empty — the actual capability surface is undocumented/unenforced.
- •http_request grants arbitrary outbound network fetches; retrieved page content is untrusted and could carry prompt-injection or instructions that steer later steps (no mitigation such as 'treat fetched content as data' is included).
- •memory_store persists the brief; the skill sensibly gates this behind explicit user request, but there is no guidance on what should not be stored (e.g., sensitive material in a research brief).
- •Minor: the skill body lists builtin tools (web_search_tool, http_request, memory_store) but no allowed-tools are declared in the manifest, so the tool surface is not machine-enforced and depends on the agent's default permissions.
- •http_request grants outbound network access; instructions only direct its use toward source gathering, with no calls to send credentials, tokens, or local file contents outward.
Examples
Research how vector databases are priced across Pinecone, Weaviate and Qdrant, and give me a short brief comparing entry-level tiers with your confidence level.Do remote workers really produce more output? Research it properly, tell me where the studies disagree, and write up a brief with sources tagged.Give me an outlined research brief on the regulatory status of AI-generated content labelling in the EU and UK, flagging anything you can't verify.Research Assistant
Description
Deep research with structured outlines, source triangulation, and an explicit confidence pass. Use this skill when the user asks an open-ended research question that benefits from multiple sources, comparison, and a written brief rather than a one-shot answer.
Tools
- name: web_search_tool kind: builtin
- name: http_request kind: builtin
- name: memory_store kind: builtin
Instructions
- Start by restating the question and proposing a brief outline; let the user confirm or amend before diving in.
- Cast a wide net: at least three independent sources before drawing a conclusion.
- Triangulate: when sources agree, say so; when they conflict, name both sides and explain the disagreement.
- Tag each claim in the brief with one of:
[primary],[secondary],[inferred]. Never present[inferred]claims as fact. - Close with a "Confidence" section: high / medium / low, with the reasons.
- Return the final brief in the reply. Store it with
memory_storeonly when the user asks to have it remembered. - If a question requires expertise outside reach (paywalled journals, proprietary data), say so explicitly rather than bluffing.
From the file to a first result
The download contains exactly the content displayed on this page. This copy may differ from the latest version in the source repository.
View the source and supporting filesThis download contains SKILL.md only. Any scripts, references, templates and dependencies are not included. Get the required files from the source, preserving their directory structure, before continuing.
1. Download the file
Download SKILL.mdFrom the root of a test project, save the file as .skill-downloads/deep-research-assistant-mvda883b/SKILL.md, or use the command below. This folder is for inspecting the skill before installation.
Command for macOS/Linux or WSL: requires a POSIX shell and curl. On Windows without WSL, use the download button and create the folder manually.
mkdir -p '.skill-downloads/deep-research-assistant-mvda883b' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/deep-research-assistant-mvda883b/download' --output '.skill-downloads/deep-research-assistant-mvda883b/SKILL.md.part' && mv '.skill-downloads/deep-research-assistant-mvda883b/SKILL.md.part' '.skill-downloads/deep-research-assistant-mvda883b/SKILL.md'2. Check prerequisites
Read SKILL.md before using it. If it references scripts or external services, inspect them and prepare the required access in your test environment.
No recognized prerequisites section in this file. This does not mean it is self-contained: libraries, tools and access requirements may be mentioned elsewhere in the content or source repository.
The platforms on this page are catalog indications. The paths below follow tool documentation; they do not prove this skill has been run in those tools. No minimum tool version is confirmed for this listing.
3. Load the skill
This file does not provide a usable name/description header for this native procedure. It remains available to download and inspect; use the manual method below.
Manual method or another tool: attach SKILL.md to a conversation, or open it in an agent that can read files. Include any required supporting resources. No universal folder is recognized by every tool.
4. Try a first task
Use a short, synthetic input: an API endpoint, a small document or an interface brief. In a new conversation, attach the file or give the agent its path, then adapt this request:
Read .skill-downloads/deep-research-assistant-mvda883b/SKILL.md and apply its instructions to this example: [describe a test input]. Identify any missing files, dependencies or access. Produce the result requested by the skill without inventing data or running unreviewed scripts.Expected result: output matching the skill instructions, or a precise list of missing prerequisites. A read confirmation alone does not validate a first use. If an external resource is unavailable, that path remains untested.
Writing Fragments
Content
Mine raw writing fragments through a relentless interview, without imposing structure. Capture ideas, quotes, and leading words for later use.
UI Copywriting & Auditing
Content
Write or audit user-facing UI copy including buttons, errors, empty states, and tooltips across the monorepo, following the design system source of truth.
Music MIDI Extraction
Content
Extracts the MIDI or score from an existing music production via the CogMusic MCP tool and confirms availability.