deep-research

deep-research is a command for coding agents from SaigonXIII/evc. It costs 0 tokens per session (660 once invoked), scanned A, original, MIT.

A command for multi-source web research that searches several angles, checks important claims across sources, and produces a cited report.

In plain words
What is it for?
Investigating competitors, markets, current events, products, or other topics that need recent information, multiple sources, synthesis, and source citations.
Why use it?
It turns a broad research question into smaller questions and combines verified findings, so the user does not have to gather and reconcile sources manually.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/saigonxiii/evc/deep-research
Clone the repo
git clone --depth 1 https://github.com/SaigonXIII/evc

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for deep-research

README.md
[![agentmods](https://agentmods.dev/badge/commands/saigonxiii/evc/deep-research.svg)](https://agentmods.dev/commands/saigonxiii/evc/deep-research)
Your own site
<a href="https://agentmods.dev/commands/saigonxiii/evc/deep-research"><img src="https://agentmods.dev/badge/commands/saigonxiii/evc/deep-research.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 660 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.00660
Opus 5 $0.00000 $0.00330
Sonnet 5 $0.00000 $0.00132
Haiku 4.5 $0.00000 $0.00066

Measured 3d ago against content hash 7bb8017e9dfd, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

deep-research scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

_templates/commands/deep-research.md · 84 lines

How it starts

The opening of the file, as written. The whole thing — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Deep Research

Multi-source research with cited reports. Searches the web, synthesizes findings, delivers structured output.

Trigger

User invokes /deep-research or says "research", "deep dive", "investigate", "competitive analysis", "what's the current state of"

Steps

1. Clarify scope

Ask clarifying questions only if the scope is genuinely ambiguous — otherwise proceed with reasonable defaults. Never block an autonomous or scheduled run on a question the context already answers.

2. Plan sub-questions

Break the topic into the sub-questions that actually decompose it. Example:

  • Topic: "Direct and indirect competitors in our category"
    • Who are the main players?
    • What's their positioning and core value prop?
    • How do they price and go to market?
    • What audience are they targeting?
    • Market size, growth, and notable shifts in the last 12 months?

3. Execute multi-source search

Search each sub-question from more than one angle — rephrase, approach by entity, by category, by time period. Scale source count to the question: a narrow factual check may need three sources; a market landscape may need dozens. Cross-verify every load-bearing claim. Prioritize: official sites, docs, reputable news > blogs > forums.

4. Deep-read key sources

WebFetch the most promising URLs and read full content. Do not rely only on search snippets for anything that matters to the conclusion.

5. Parallel research (for broad topics)

Delegate independent sub-questions to parallel subagents — decide the count and split by what the topic needs. Each agent searches, reads sources, returns findings. Main session synthesizes.

6. Synthesize report

# [Topic]: Research Report
*Generated: [date] | Sources: [N] | Confidence: [High/Medium/Low]*

## Executive Summary
[3-5 sentences]

## 1. [First Theme]
[Findings with inline citations]
- Key point ([Source Name](url))

## 2. [Second Theme]
...

## Key Takeaways
- [Actionable insight 1]
- [Actionable insight 2]
- [Actionable insight 3]

## Sources
1. [Title](url) — [one-line summary]

## Methodology
Searched [N] queries. Analyzed [M] sources.

Read the full file on GitHub · 84 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 84 lines · 0 tokens per session scan A 7bb8017e9dfd

Subscribe to this mod's changes

deep-research is a command published in the GitHub repository SaigonXIII/evc (56 stars, last pushed 2mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 660 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.