deep-research

A research workflow that uses gpt2agent to request either a shorter or more extensive research report, with citations retained.

In plain words
What is it for?
Use it to run research queries from text or files and save the resulting cited reports.
Why use it?
It removes the need to gather and assemble research findings manually, while keeping source references with the report.

Skill for Claude CodeCodex

Part of the gpt2agent plugin — 2 skills shipped together

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 skills/robotlearning123/gpt2agent/deep-research
Any agent
npx skills add robotlearning123/gpt2agent --skill deep-research
Clone the repo
git clone --depth 1 https://github.com/robotlearning123/gpt2agent

Made for: Claude Code, Codex.

Or install gpt2agent, the plugin that ships this one along with the rest of its 2 skills.

Per session 163 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,057 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 $0.00163 $0.02057
Opus 5 $0.00081 $0.01028
Sonnet 5 $0.00033 $0.00411
Haiku 4.5 $0.00016 $0.00206

Measured 3d ago against content hash 39ebc3c77c1c, 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.

The scan reads SKILL.md. This mod also ships 3 executable files (bin/deep_research.py, bin/quota.sh, bin/run.sh), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

gpt2agent/skills/deep-research/SKILL.md · 170 lines

How it starts

The opening of the file, as written. The whole thing — 170 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/deep-research — ChatGPT Pro Deep Research

Calls gpt2agent's deep_research / deep_research_heavy directly via pipx Python (bypasses MCP — works even before Claude Code session restart).

Preconditions (check once)

command -v gpt2agent >/dev/null || \
  echo "gpt2agent not installed; run: pipx install gpt2agent"
test -f "${CODEX_HOME:-$HOME/.codex}/auth.json" || test -f "$HOME/.gpt2agent/token.json" || \
  echo "ChatGPT token missing; run: codex login or gpt2agent setup"
~/.claude/skills/deep-research/bin/quota.sh   # prints remaining DR quota

Usage

~/.claude/skills/deep-research/bin/run.sh [--heavy] [-o OUT_DIR] "<query>"
  • Default mode: light (deep_research, ~1 min, citations included).
  • --heavy: deep_research_heavy (5-30 min, gpt-5-5-pro + connector). The connector renders an embedded-UI widget; the report is recovered from the hidden widget state (widget_state.report_message) via ?include_visually_hidden_messages=true&include_widget_state=true — see "Heavy DR retrieval" below.
  • -o OUT_DIR: output directory (default: a unique ./research/dr-YYYYMMDD-HHMMSS-*/ directory).
  • Query can be inline string, - for stdin, or @file.md to read from file.

The script writes:

  • report.md — final report (reconstructed for heavy mode)
  • events.jsonl — all raw SSE events (for debugging / re-extraction)
  • status.txt — START / DONE / INCOMPLETE / ERROR with elapsed seconds + event counts
  • meta.json — server metadata (model slug, request id, etc.)

The run directory is restricted to mode 0700 and its artifacts to 0600 on POSIX systems because queries, reports, and metadata may be sensitive.

When to invoke

Situation Mode
Quick factual question with citations light
Literature review, market scan, technical decision matrix light
Big strategic question (>5 questions, want 10+ KB report) heavy
Question that might yield 50+ sources heavy

Skip this skill for: pure code questions, debugging, tasks the user explicitly wants you to handle locally, anything covered by context7 (library docs) or local files.

Read the full file on GitHub · 170 lines

Files

What ships with it

3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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 · 170 lines · 163 tokens per session scan A 39ebc3c77c1c

Subscribe to this mod's changes

deep-research is a skill published in the GitHub repository robotlearning123/gpt2agent (47 stars, last pushed 1mo ago), licensed MIT. It adds 163 tokens to every session and 2,057 once invoked, about $0.0008 per session on Opus 5. 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.

Related

Other skills, from other repositories

search

Search 2500+ curated ChatGPT and LLM open-source repositories. Use when the user asks to find tools, libraries, or repos related to ChatGPT, LLMs, RAG, agents, langchain, NLP, AI development, or any open-source AI tooling.

taishi-i/awesome-ChatGPT-repositories · 57 tokens

pptx

Create and validate Microsoft PowerPoint presentations (.pptx), including structured slide decks, tables, workflows, metadata, and reproducible generation scripts. Use for presentation, slides, PowerPoint, PPT, or PPTX creation and verification tasks.

agents-flex/agents-flex · 50 tokens

chatgpt-app-builder

Guide developers through creating and updating ChatGPT apps. Covers the full lifecycle: brainstorming ideas against UX guidelines, bootstrapping projects, implementing tools/views, debugging, running dev servers, deploying and connecting apps to ChatGPT. Use when a user wants to create or update a ChatGPT app / MCP…

alpic-ai/skybridge · 78 tokens

ai-tutor

Use when user asks to explain, break down, or help understand technical concepts (AI, ML, or other technical topics). Makes complex ideas accessible through plain English and narrative structure. Use the provided scripts to transcribe videos.

agents-flex/agents-flex · 48 tokens

add-app-language

Use when adding a new supported application UI language or locale family to all-api-hub, including i18next resources, language selection and normalization, browser extension manifest locales, date localization or natural-language date parsing, documentation-link fallback, locale completeness tests, and validation.

qixing-jk/all-api-hub · 57 tokens

sponsor-catalog

Use when adding, updating, removing, reordering, or auditing sponsors, affiliate links, sponsor copy, logos, public sponsor-catalog JSON files, README sponsor sections, or documentation-site sponsor listings in all-api-hub.

qixing-jk/all-api-hub · 50 tokens