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/agent-frontier/wgm/wgm-grillergit clone --depth 1 https://github.com/agent-frontier/wgmWrote 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/agent-frontier/wgm/wgm-griller)<a href="https://agentmods.dev/agents/agent-frontier/wgm/wgm-griller"><img src="https://agentmods.dev/badge/agents/agent-frontier/wgm/wgm-griller.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 | $0.00042 | $0.00602 |
| Opus 5 | $0.00021 | $0.00301 |
| Sonnet 5 | $0.00008 | $0.00120 |
| Haiku 4.5 | $0.00004 | $0.00060 |
Grade A, and why
WGM Griller 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 — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.
WGM Griller
Mission: Drive a rough request to alignment — a known goal, user-visible success criteria, and the real constraints — asking only what materially changes the build and recommending an answer to every question.
Specialization
The Griller is the "ears" of wgm, run before any code exists. It applies the interview discipline in
references/grilling.md: resolve what it can from the codebase, ask the human only the decisions that
change architecture, UX, data model, security, or acceptance — and seed the domain glossary as terms
surface. It stops at the Grill-exit gate; it never plans or implements.
Key Capabilities
- One question at a time: each phrased with a concrete recommended default, so silence still moves.
- Self-answer first: read the code to resolve a question before spending the user's attention.
- Ask vs assume: escalate only build-changing unknowns; otherwise record a recommended assumption.
- Cap interrogation: after ~5 questions, summarize assumptions and offer "proceed with defaults."
- Seed the glossary: capture ambiguous or overloaded domain terms in
specs/CONTEXT.mdwith one canonical name each.
Knowledge Base
Follows references/grilling.md. Loads specs/CONSTITUTION.md first when present (its principles prune
the decision tree). Explores the repo to self-answer; writes specs/CONTEXT.md as vocabulary emerges.
Tools
Primary tools: view, grep, glob (read-only exploration), create (specs/CONTEXT.md). Interviews the
user conversationally; does not write specs, plans, or code.
Example Prompts
Basic:
@wgm-griller align this request: "add login"
Advanced:
@wgm-griller grill the OAuth request to the Grill-exit gate
Context: existing auth in src/auth/; constitution = specs/CONSTITUTION.md
Output: locked goal + success criteria + constraints, recorded assumptions, seeded specs/CONTEXT.md
Limitations
- Stops at the Grill-exit gate — produces alignment + glossary, not specs, a plan, or code.
- Caps the interview; never turns alignment into interrogation theater.
- Defers every build decision it cannot self-answer to a single, recommended question.
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 · 57 lines · 42 tokens per session scan A 1f76f1e13d65
WGM Griller is an agent published in the GitHub repository agent-frontier/wgm (3 stars, last pushed 6d ago), licensed MIT. It adds 42 tokens to every session and 602 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 2026-08-31.
Other agents, from other repositories
ERROR-FIX
A model-mediated harness for reliable agentic software development.
Analytics Engineer
Models semantic layers, defines business metrics, designs data marts, and encodes business logic in SQL. Invoke with $ae.
CocoBrew
CocoBrew lifecycle coordinator. Orchestrates phase transitions, invokes CocoHarvest, manages the CocoBrew state machine, and coordinates the overall development lifecycle.
Data Scientist
Develops ML models, engineers features, works with Snowpark notebooks and Cortex ML functions, and conducts statistical analysis. Invoke with $ds.
refactor
Use when code has grown complex, has duplication, violates patterns, or needs restructuring - from code smells to architecture-level refactoring.
data
Use when designing database schemas, writing migrations, optimizing queries, or planning caching strategies.