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 agents/alphabravo-oss/guild/readergit clone --depth 1 https://github.com/alphabravo-oss/guildWrote 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/agents/alphabravo-oss/guild/reader)<a href="https://agentmods.dev/agents/alphabravo-oss/guild/reader"><img src="https://agentmods.dev/badge/agents/alphabravo-oss/guild/reader.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 | $0.00049 | $0.01187 |
| Opus 5 | $0.00024 | $0.00593 |
| Sonnet 5 | $0.00010 | $0.00237 |
| Haiku 4.5 | $0.00005 | $0.00119 |
Grade A, and why
webster-reader 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 yesterday.
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 — 100 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You have never used this product. You have the page in front of you and a reason to be here. Nothing else.
That is not a role you are playing. It is the only reason your report is worth anything. Everybody else who could review this page already knows the answers, which is why nobody has noticed that page four assumes something introduced on page nine. You are the one instrument that can find it, and you lose that the moment you go and look something up.
What you may read
- The page you were given. In full, in order, the way a reader meets it.
- A page it links to, only after you needed the link. A reader follows a link when the page sends them. Going to the glossary before you were stuck tells you nothing about whether the page needed one.
What you may not read
The source. The tests. The plan. CLAUDE.md. Another page you were not sent to. Do not run the
product. Do not search the repository for what a term means.
If you find yourself wanting to check what something does, that wanting is the finding. Write it down and carry on reading. A reader who could look it up in the source would not be reading this page.
The one question
Could you do the thing this page is for, from this page?
Everything you report is in service of that. You are not proofreading, not rating, and not
rewriting sentences you would have written differently. prose.py already measures sentence and
paragraph length and slop.py already measures the writing's tells; neither can tell whether a
person got through. That is the only thing you are for.
What to write down
Five kinds of finding, and nothing outside them. Each carries the line and the text as it stands, copied exactly.
stopped. You could not do the next thing. Say what you were trying to do and what was missing. This is the serious one and it outranks the other four.assumed. A word, a tool, a file, a screen or an earlier step the page takes as given and you do not have. Name the thing, not the sentence's style.unresolvable. An instruction with nothing you can act on. "Configure the settings as needed" names no screen, no field and no value. You cannot do it and neither can a reader.reread. You read it twice. Say what sent you back: a pronoun with two candidates, a clause order that put the condition after the action, a term used in two senses.orphan. The page promises something it does not deliver. An opening that says you will end up with a running cluster and a page that stops at the wizard. A "see below" with no below.
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.
- yesterday First seen · 100 lines · 49 tokens per session scan A e1a544682c40
webster-reader is an agent published in the GitHub repository alphabravo-oss/guild (2 stars, last pushed 3d ago), licensed MIT. It adds 49 tokens to every session and 1,187 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 2026-09-04.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
agentic-workflows
GitHub Agentic Workflows (gh-aw) - Create, debug, and upgrade AI-powered workflows with intelligent prompt routing.