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 skills/mrobinson2/azureagentforge/web-researchnpx skills add mrobinson2/AzureAgentForge --skill web-researchgit clone --depth 1 https://github.com/mrobinson2/AzureAgentForgeWrote 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/mrobinson2/azureagentforge/web-research)<a href="https://agentmods.dev/skills/mrobinson2/azureagentforge/web-research"><img src="https://agentmods.dev/badge/skills/mrobinson2/azureagentforge/web-research.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 | $0.00110 | $0.00844 |
| Opus 5 | $0.00055 | $0.00422 |
| Sonnet 5 | $0.00022 | $0.00169 |
| Haiku 4.5 | $0.00011 | $0.00084 |
Grade A, and why
web-research scanned grade A with 1 finding 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
| **Read a page** — get the actual content of a URL | `web-read` | raw `curl` + strip | `web-read` returns clean Markdown; `curl` returns raw HTML. | How it starts
The opening of the file, as written. The whole thing — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Web research: pick the right backend, with fallbacks
The platform exposes several research CLIs on $PATH. Use the one that fits the
job; if it fails or returns empty, fall back down the chain rather than giving up
(or fabricating).
Routing table
| Need | Preferred | Fallback | Notes |
|---|---|---|---|
| Search — conceptual / "things like X" / exploratory | exa-search |
brave-search → ddgs |
Exa is semantic; finds relevant pages without exact words. |
| Search — fresh / news / exact term | brave-search |
exa-search → ddgs |
Brave accepts cloud IPs (DuckDuckGo blocks them). |
| Read a page — get the actual content of a URL | web-read |
raw curl + strip |
web-read returns clean Markdown; curl returns raw HTML. |
| Video — transcript of a YouTube/Vimeo/etc. link | video-transcript |
(Whisper, auto) | yt-dlp captions; Whisper fallback runs automatically if GROQ_API_KEY is set. |
All search tools emit the same JSON shape ([{title, href, body}]), so you can
swap one for another without changing how you parse the result. When a question
has both an exploratory and a freshness angle, run both search backends and
merge — they surface different pages.
Fallback rule
If your preferred tool returns an empty result or a non-zero exit, try the fallback before concluding "nothing found". An empty result from one backend is not evidence that the information doesn't exist — it's often that backend being blocked, rate-limited, or unconfigured. Only after the preferred tool and its fallback come up empty should you treat the query as genuinely unanswerable.
Health check — research-doctor
Before a research-heavy task (or when a tool behaves oddly), check which backends are actually up:
research-doctor # human-readable: OK / SKIP / DOWN per backend
research-doctor --json # [{name, ok, status, detail}] for scripting
- OK — the backend responded.
- SKIP — not configured here (e.g. no API key); not an error, just unavailable.
- DOWN — configured but failing; route around it and, if persistent, flag it
to Infrastructure.
research-doctorexits non-zero when any backend is DOWN.
What ships with it
1 file 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.
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 · 67 lines · 110 tokens per session scan A 5d95af87f866
web-research is a skill published in the GitHub repository mrobinson2/AzureAgentForge (21 stars, last pushed 14d ago), licensed MIT. It adds 110 tokens to every session and 844 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
Cloud Security & Container Hardening
AWS/Azure/GCP security auditing, container and Kubernetes hardening, Infrastructure as Code scanning, and cloud compliance assessment.
terraform-expert
Expert-level Terraform infrastructure as code, modules, state management, and production best practices.
azure-expert
Expert-level Microsoft Azure cloud platform, services, and architecture.
tomllm-format
.tomllm / .tomllmd = valid TOML + enriched # comment conventions for LLM context.
terraform-iac
Terraform and OpenTofu infrastructure as code best practices - generate HCL configurations, module patterns, state management, CI/CD workflows, and cloud provider templates for AWS, GCP, Azure.
terraform-infrastructure-as-code
Comprehensive Terraform Infrastructure as Code skill covering resources, modules, state management, workspaces, providers, and advanced patterns for cloud-agnostic infrastructure deployment.