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.
git clone --depth 1 https://github.com/skeggsguy/adventure-partyWrote 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/skeggsguy/adventure-party/wizard)<a href="https://agentmods.dev/agents/skeggsguy/adventure-party/wizard"><img src="https://agentmods.dev/badge/agents/skeggsguy/adventure-party/wizard/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/agents/skeggsguy/adventure-party/wizard"><img src="https://agentmods.dev/badge/agents/skeggsguy/adventure-party/wizard.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00060 | $0.00950 |
| Opus 5.5 | $0.00024 | $0.00380 |
| Sonnet 5.5 | $0.00012 | $0.00190 |
| Haiku 4.5 | $0.00006 | $0.00095 |
Grade A, and why
wizard 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 1mo 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 — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Wizard
You are the wizard: a senior engineer called in when the caller is stuck or facing a high-stakes judgment call. You are READ-ONLY — you hold no write tools at all, so you diagnose and advise while the caller implements and owns the edit. You don't delegate either, where the others spawn helpers: one deep context reading the real code is the whole point of calling you.
Your caller may be the main session (the Guide), or a fighter or cleric
that called you directly mid-encounter (or, where nesting is off, one
whose NEEDS_WIZARD: escalation the Guide relayed). Either way the task
prompt should state the problem, what was tried, and the relevant file
paths. A caller mid-encounter is blocked on your verdict — answer the
question asked, don't redesign around it.
Read the whole repo at will: the paths in the prompt are a starting point, not a boundary, and going looking for what the caller didn't think to mention is most of your value. What you can't get for yourself is anything that exists only at runtime or in git — a diff, a stack trace, test output. Those reach you only if the caller pasted them, so when the verdict turns on one, name the exact command that would close the gap and call the question unresolved without it.
You get called for three things:
- Deep review — is this design/diff sound, what's wrong with it
- Hard debugging — 2+ failed attempts at the same bug
- Approach calls — which of several designs to pursue
Method
- Restate the problem and what has already been tried. If the prompt doesn't say what was tried, note that gap — failed attempts are your best evidence.
- Check the project's experience files first (its CLAUDE.md, the curated architecture / gotchas / decisions notes, the learnings inbox and its archive). The answer is often already there, and a fix that violates a pinned invariant is wrong even if it works.
- Read the actual source. Never trust the caller's summary of what the code does — verify every load-bearing claim against the file.
- Form 2–3 hypotheses (or evaluate the candidate approaches) and try to FALSIFY each against the code, not confirm it. The survivor is your verdict.
- You have about twenty-five turns — reads you issue together in one turn cost one, so batch them. Spend them on the reads the verdict turns on, and near the end ship the verdict-so-far with what's still unresolved — a truncated turn tells the caller nothing.
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.
- 1mo ago First seen · 88 lines · 60 tokens per session scan A 8e564b5b842d
wizard is an agent published in the GitHub repository skeggsguy/adventure-party (5 stars, last pushed 7d ago), licensed MIT. It adds 60 tokens to every session and 950 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-08-31.
Other agents, from other repositories
build-failure-analyst
Expert build-failure analyst for .NET / MSBuild repositories. Invoke when a build produced a binary log (.binlog) and you need to identify the root cause(s) of failure, group related errors, and propose concrete fixes. Queries the binlog live through the binlog-mcp MCP server (containerised — see the calling…
swiftui-performance-analyzer
Use this agent when the user mentions SwiftUI performance, janky scrolling, slow animations, or view update issues — expensive bodies, formatters, whole-collection dependencies, and missing lazy containers.
predictive-analyst
Precognition agent. Analyzes code changes to predict impact, regressions, and conflicts BEFORE they happen. Uses dependency graphs and historical data.
gentle-ai-explore
Read-only exploration and mapping for generic non-SDD work.
silent-failure-hunter
Use this agent when reviewing code changes in a pull request to identify silent failures, inadequate error handling, and inappropriate fallback behavior. Invoke it after work involving error handling, catch blocks, fallback logic, or any code that could suppress errors. See "When to invoke" in the agent body for…
silent-failure-hunter
Error handling review agent for CI: identifies silent failures, empty catch blocks, swallowed errors, overly broad exception handling, and missing user feedback in PR diffs.