Pydantic Deep Agents is a self-hosted terminal AI assistant and Python framework for building coding, research, and other AI agents. It gives agents tools such as file access, shell commands, planning, memory, sub-agents, sandboxed execution, and MCP connections, and supports different models.
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 vstorm-co/pydantic-deepagents --skill research-methodologygit clone --depth 1 https://github.com/vstorm-co/pydantic-deepagentsWrote 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/vstorm-co/pydantic-deepagents/research-methodology)<a href="https://agentmods.dev/skills/vstorm-co/pydantic-deepagents/research-methodology"><img src="https://agentmods.dev/badge/skills/vstorm-co/pydantic-deepagents/research-methodology/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/vstorm-co/pydantic-deepagents/research-methodology"><img src="https://agentmods.dev/badge/skills/vstorm-co/pydantic-deepagents/research-methodology.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00016 | $0.00515 |
| Opus 5 | $0.00008 | $0.00258 |
| Sonnet 5 | $0.00003 | $0.00103 |
| Haiku 4.5 | $0.00002 | $0.00052 |
Grade A, and why
research-methodology 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 9d 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Research Methodology Guide
Search Strategy
Phase 1: Broad Discovery
- Start with general queries to understand the landscape
- Use different phrasings for the same concept
- Note key terminology, authors, and organizations
Phase 2: Focused Deep-Dive
- Search for specific claims, statistics, or technical details
- Target authoritative sources identified in Phase 1
- Use exact phrases in quotes for precision
Phase 3: Verification
- Cross-reference key claims across multiple sources
- Search for counter-arguments or contradictions
- Check publication dates for recency
Source Evaluation
Reliability Hierarchy
- Academic papers (peer-reviewed journals, arXiv preprints)
- Official documentation (government, organization, project docs)
- Reputable news (established outlets with editorial standards)
- Expert blog posts (known authors with credentials)
- Community forums (Stack Overflow, Reddit — use cautiously)
Evaluation Checklist
- Authority: Who wrote it? What are their credentials?
- Currency: When was it published? Is it still relevant?
- Coverage: Does it address the topic comprehensively?
- Accuracy: Can claims be verified elsewhere?
- Objectivity: Is there obvious bias or commercial interest?
Note-Taking Best Practices
Structure Each Note File
# [Sub-topic Title]
## Key Findings
- Finding 1 [SOURCE: url, accessed YYYY-MM-DD] [HIGH confidence]
- Finding 2 [SOURCE: url, accessed YYYY-MM-DD] [MEDIUM confidence]
## Contradictions
- Source A says X, but Source B says Y
## Gaps
- Could not find reliable data on Z
Confidence Levels
- [HIGH]: Multiple authoritative sources agree
- [MEDIUM]: Single authoritative source, or multiple less-reliable sources agree
- [LOW]: Single non-authoritative source, or conflicting information
Common Pitfalls
- Don't rely on a single source for important claims
- Check if "recent" articles cite outdated data
- Be wary of sources that don't cite their own sources
- Distinguish between correlation and causation
- Note when sample sizes are small or studies are preliminary
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.
- 9d ago First seen · 70 lines · 16 tokens per session scan A 5e5287c1dad1
research-methodology is a skill published in the GitHub repository vstorm-co/pydantic-deepagents (1,059 stars, last pushed 18d ago), licensed MIT. It adds 16 tokens to every session and 515 once invoked, about $0.0001 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
pydantic-ai
Build production-ready AI agents with PydanticAI — type-safe tool use, structured outputs, dependency injection, and multi-model support.
hive.slack-notifications-setup
Set up a Slack notification channel (Sentinel) for a colony by driving the browser — reuse or create the "Hive Sentinel" Slack app from a JSON manifest, install it, capture the bot + app tokens, create/select the channel via the Slack API, and turn Sentinel on so the colony can ping the user on Slack and accept…
hive.chart-creation-foundations
Required reading whenever any chart tool is available. Teaches the one-tool embedding contract (call chartrender → live chart appears in chat AND a downloadable PNG lands in the queen session dir), the ECharts (data viz) vs Mermaid (structural diagrams) decision, the BI/financial-grade aesthetic baseline (no…
browser-edge-cases
SOP for debugging browser automation failures on complex websites. Use when browser tools fail on specific sites like LinkedIn, Twitter/X, SPAs, or sites with Shadow DOM.
hive.pdf
Read, write, merge, split, rotate, watermark, encrypt, and OCR PDF files using Python (pypdf, pdfplumber, reportlab, pypdfium2) and command-line tools (poppler-utils, qpdf). Use when the user asks to extract text/tables/images from a PDF, create or modify a PDF, combine or split PDFs, OCR a scanned PDF…
hive.note-taking
Maintain a free-form scratchpad of decisions, extracted values, and open questions so context pruning doesn't lose anything you still need.