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.
npx skills add tonyazhuuki/deep-research-skill --skill researchgit clone --depth 1 https://github.com/tonyazhuuki/deep-research-skillWrote 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.
[](https://agentmods.dev/skills/tonyazhuuki/deep-research-skill/research)<a href="https://agentmods.dev/skills/tonyazhuuki/deep-research-skill/research"><img src="https://agentmods.dev/badge/skills/tonyazhuuki/deep-research-skill/research/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/tonyazhuuki/deep-research-skill/research"><img src="https://agentmods.dev/badge/skills/tonyazhuuki/deep-research-skill/research.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00054 | $0.02394 |
| Opus 5 | $0.00027 | $0.01197 |
| Sonnet 5 | $0.00011 | $0.00479 |
| Haiku 4.5 | $0.00005 | $0.00239 |
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 10d 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.
How it starts
The opening of the file, as written. The whole thing — 201 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Deep Research Skill
A structured multi-agent research pipeline that turns any topic into a comprehensive, fact-checked, bilingual (EN+RU) synthesis with actionable recommendations. Built on Eric Jang's iterative methodology from "As Rocks May Think".
How It Works
The skill orchestrates 10-19 specialized AI agents across 3 mandatory cycles:
Cycle 1: Broad Search → 4-5 parallel SCOUTs explore the landscape
Quality Gates → CRITIC + METHODOLOGIST cross-check findings
Reflection 1 → Identify gaps, generate competing hypotheses
Cycle 2: Deep Dives → 2-3 targeted agents test hypotheses + stress-test questions
Iterative Deep. → Auto-resolve CONTESTED claims (WEAK → follow-up DD)
Reflection 2 → Convergence analysis, hypothesis verdicts
Cycle 3: Execute → Python scripts (analysis, models, visualizations)
Synthesize → SYNTHESIZER creates integrated document
Verify → FACT-CHECKER + CITATION_VERIFIER + DOMAIN_REVIEWER
Apply → ACTION MAPPER updates user's protocols/goals
Research Modes
| Mode | Output | When to use |
|---|---|---|
| personalized (default) | synthesis.md |
Specific question for your context |
| consensus | consensus_reference.md |
Building knowledge base (population-level truth) |
| consensus+interactions | consensus + interaction_map.md |
Cross-effects matter |
| full | All three documents | Deep investigation |
Agent Roles
| Role | Count | Purpose |
|---|---|---|
| SCOUT | 4-5 | Broad literature search, each with unique reasoning style |
| CRITIC | 1 | Cross-stream contradictions, bias audit, weak evidence |
| METHODOLOGIST | 1 | Methodological quality — domain-specific evidence hierarchy (GRADE for health, forecast audit for macro, reproducibility for science) |
| DEEP DIVER | 2-3+ | Hypothesis testing + domain stress-test questions. Auto-spawns on CONTESTED claims (iterative deepening) |
| SYNTHESIZER | 1 | Integration across all sources into coherent document |
| INTERACTION MAPPER | 1 | Cross-domain interactions that change recommendations |
| DOMAIN_REVIEWER | 1 | Domain-specific review: MEDICAL (health), MACRO (markets), MARKET (company), METHODOLOGY (science) |
| FACT-CHECKER | 1 | Top-15 numerical claims verification |
| CITATION_VERIFIER | script | Python API check against Semantic Scholar/PubMed/CrossRef |
| TEMPORAL DIFF | 0-1 | Compares new consensus with previous version (UPDATE mode only) |
| ACTION MAPPER | 1 | Converts findings into TODO blocks in user's files |
What ships with it
29 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.
- adapters/genome.md 7.6 KB
- adapters/README.md 1.9 KB
- agents.md 14 KB
- context_template.md 9.5 KB
- cycle1.md 13 KB
- cycle2.md 3.5 KB
- cycle3.md 14 KB
- domains/company.md 21 KB
- domains/creative.md 21 KB
- domains/health_databases.md 8.7 KB
- domains/health.md 20 KB
- domains/macro.md 17 KB
- domains/marketing.md 45 KB
- domains/science.md 20 KB
- examples/example_output_tree.txt 3.0 KB
- examples/example_synthesis_excerpt.md 1.8 KB
- finalize.md 5.6 KB
- INSTALL.md 8.7 KB
- OVERVIEW.md 20 KB
- prompts.md 85 KB
- README.md 1.5 KB
- templates/README.md 2.1 KB
- templates/study_card_company.yaml 3.4 KB
- templates/study_card_creative.yaml 3.2 KB
- templates/study_card_health.yaml 6.9 KB
- templates/study_card_macro.yaml 3.2 KB
- templates/study_card_marketing.yaml 6.6 KB
- templates/study_card_science.yaml 3.4 KB
- templates/synthesis_longread.md 5.9 KB
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.
- 10d ago First seen · 201 lines · 54 tokens per session scan A e8a11509a9d9
deep-research is a skill published in the GitHub repository tonyazhuuki/deep-research-skill (31 stars, last pushed 1mo ago), licensed MIT. It adds 54 tokens to every session and 2,394 once invoked, about $0.0003 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.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…