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 agentmods add commands/a-ariff/ariff-claude-plugins/deep-searchgit clone --depth 1 https://github.com/a-ariff/ariff-claude-pluginsWrote 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/commands/a-ariff/ariff-claude-plugins/deep-search)<a href="https://agentmods.dev/commands/a-ariff/ariff-claude-plugins/deep-search"><img src="https://agentmods.dev/badge/commands/a-ariff/ariff-claude-plugins/deep-search.svg" alt="Measured on agentmods" 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 | $0.00000 | $0.00776 |
| Opus 5 | $0.00000 | $0.00388 |
| Sonnet 5 | $0.00000 | $0.00155 |
| Haiku 4.5 | $0.00000 | $0.00078 |
Grade A, and why
deep-search 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 5d 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 — 111 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Deep Search Command
Intelligent search across multiple sources using semantic vendor detection.
Usage
/deep-search <query>
How It Works
- Analyze Query - Detect topics, technologies, vendors
- Route to Sources - Select relevant sources based on topic
- Execute Search - Query each source
- Synthesize - Combine and summarize results
Source Routing
Always Searched
- GitHub - Code examples, implementations, issues
Semantic Routing
| Topic Detected | Sources Added |
|---|---|
| Claude, Anthropic | docs.anthropic.com, Claude guides |
| Azure, Microsoft | Microsoft Learn, Azure docs |
| Intune, Endpoint | Microsoft Intune docs |
| Python | Python docs, PyPI |
| JavaScript, Node | MDN, npm |
| Canvas LMS | Canvas API docs |
| Terraform | Terraform Registry, HashiCorp docs |
| Kubernetes, K8s | kubernetes.io |
| Docker | Docker docs |
| AWS | AWS docs |
| Reddit discussion | reddit.com (via firecrawl) |
Example Searches
/deep-search claude code hooks best practices
→ GitHub (claude-code), docs.anthropic.com
/deep-search intune windows 11 compliance policy
→ GitHub, Microsoft Learn, Intune docs
/deep-search python async best practices reddit
→ GitHub, Python docs, Reddit
/deep-search terraform azure container apps
→ GitHub, Terraform Registry, Azure docs
Implementation
Uses firecrawl-mcp for web search and scraping:
# Search with firecrawl
mcp__firecrawl-mcp__firecrawl_search(
query="claude hooks best practices site:github.com OR site:docs.anthropic.com",
limit=10
)
# For Reddit
mcp__firecrawl-mcp__firecrawl_search(
query="site:reddit.com {query}",
limit=5
)
Manual Deep Search Steps
When using without command:
-
GitHub Search:
Use github_repo or grep_search for code -
Web Search:
Use mcp__firecrawl-mcp__firecrawl_search with site: operators -
Fetch Full Content:
Use mcp__firecrawl-mcp__firecrawl_scrape for specific URLs
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.
- 5d ago First seen · 111 lines · 0 tokens per session scan A 804ff8644a01
deep-search is a command published in the GitHub repository a-ariff/ariff-claude-plugins (14 stars, last pushed 5mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 776 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.
Other commands, from other repositories
run-autoresearch
Run an autonomous experiment loop via the Autoresearch Orchestrator agent.
taskrunner-add
Add a task to the autonomous queue.
taskrunner
Run pending tasks from the queue, show status, or execute a specific number of tasks.
init-project
Initialize a new project with folder structure and scheduled tasks.
taskrunner-list
Show the current task queue.
sprint-report
Generate weekly sprint progress report.