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 sliamh11/Deus --skill deep-researchgit clone --depth 1 https://github.com/sliamh11/DeusWrote 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/sliamh11/deus/deep-research)<a href="https://agentmods.dev/skills/sliamh11/deus/deep-research"><img src="https://agentmods.dev/badge/skills/sliamh11/deus/deep-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/sliamh11/deus/deep-research"><img src="https://agentmods.dev/badge/skills/sliamh11/deus/deep-research.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 44 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00047 | $0.02878 |
| Opus 5 | $0.00023 | $0.01439 |
| Sonnet 5 | $0.00009 | $0.00576 |
| Haiku 4.5 | $0.00005 | $0.00288 |
Grade A, and why
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 6d 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 — 321 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Deep Research Pipeline
Host-side only (Claude Code). See docs/agent-agnostic-debt.md AAG-012 for backend parity status.
A 4-stage research pipeline. Classifies research depth, clarifies scope when ambiguous, then routes to shallow or deep retrieval with structured citation-backed output.
Design pattern: Mediator. This skill is the central coordinator. Research scouts and brainstormer are independent peers invoked through the Agent tool without cross-coupling. Neither agent knows about the other or about the pipeline stages.
Composes existing infrastructure — does NOT duplicate it:
- Research scout agents for evidence-classified source finding (deep path)
- brainstormer agent for creative synthesis (when the user wants ideas, not just facts)
- Parallel AI MCP for web search when available (graceful fallback to WebSearch)
- Vault memory for prior decisions and research
- NotebookLM for querying existing notebooks when relevant
Instructions
When the user invokes /deep-research <topic> or a trigger phrase matches:
Stage 1: Classify intent
Read the query and classify into one of:
| Depth | Signal | Example |
|---|---|---|
| SHALLOW | Single fact, narrow scope, recent event, quick comparison | "What's the latest on X?", "Compare A vs B briefly" |
| DEEP | Multi-source synthesis, historical overview, regulatory landscape, decision brief, "research X" | "Research the regulatory landscape for Y", "Produce a brief on Z" |
| CREATIVE | Solution design, brainstorming, "how could we", exploration of alternatives | "How could we improve X?", "What are creative approaches to Y?" |
State the classification and proceed. Do NOT ask the user to confirm depth — only ask if the topic itself is ambiguous (Stage 2).
Stage 2: Clarify scope (conditional)
Skip this stage if the query is specific enough to research directly. Most queries are.
Only use AskUserQuestion when genuinely ambiguous — when researching the wrong scope would waste significant effort:
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.
- 6d ago First seen · 321 lines · 47 tokens per session scan A 0d459bb9eaa0
research is a skill published in the GitHub repository sliamh11/Deus (51 stars, last pushed 4d ago), licensed MIT. It adds 47 tokens to every session and 2,878 once invoked, about $0.0002 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 2026-09-03.
Other skills, from other repositories
memcord
Privacy-first, self-hosted chat memory for OpenClaw — save and recall conversation history across sessions without any cloud dependency.
roadmap-build
Build or repair a roadmap that agents can actually execute, not just read. Produces a single validated YAML source of truth, generated views, definitions of done with evidence kinds, a task pool linked on two axes, an idea intake queue, and three standing checks that keep it honest. Use when asked to "build a…
Research for Datacore
Automated research — source processing, podcast generation, and knowledge extraction.
GTD for Datacore
Getting Things Done — task capture, inbox processing, and org-mode management.
Outbox for Datacore
Content routing — archive to server, search archived content.
research-status
Check research queue counts, recent podcasts, knowledge base growth, and Readwise items.