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/stainless-code/codemap/grillingnpx skills add stainless-code/codemap --skill grillinggit clone --depth 1 https://github.com/stainless-code/codemapWhat 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 | $0.00024 | $0.00816 |
| Opus 5 | $0.00012 | $0.00408 |
| Sonnet 5 | $0.00005 | $0.00163 |
| Haiku 4.5 | $0.00002 | $0.00082 |
Grade A, and why
grilling 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 — 50 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Grilling
Interview relentlessly about every aspect of a plan until shared understanding. Walk down each branch of the decision tree, resolving dependencies between decisions one-by-one. For each question, provide a recommended answer. Ask one question at a time.
Rules
- One question at a time. Wait for the user's answer before posing the next. Multi-question batches collapse the decision tree into a wall of text.
- Always recommend. Every question carries your strongest read on the answer + a one-line "why". The user wants a sparring partner, not a menu.
- Explore before asking. If the answer lives in source, use
codemap queryfor structural facts, thenRead/Globbefore asking. Questions that the codebase already answers waste turns and signal weak prep. - Walk dependencies, not breadth. Resolve one branch fully before moving sideways. If Q3's answer makes Q5 moot, kill Q5; if Q3 forks into two new sub-questions, ask the sub-questions before returning to Q4.
- Push back on weak answers. If the user's answer leaves the design ambiguous or contradicts something earlier, surface the conflict and re-pose. Don't accept "we'll see" — that's a Q for next time, not closure.
- Stop when the design is operational. "Operational" means a competent contributor could open a PR from the resulting plan without further design questions. Not before, not after.
Interaction shape
Agent: <Q1>. My read: <recommended answer>. Why: <1-line reason>.
User: <answer / pushback>
Agent: <Q2 — derived from Q1's answer>. My read: …
User: …
Agent: That closes the design tree for me. Want me to <draft the plan / open a PR / start coding>?
Don't number the questions in chat — the conversation IS the order. Number them only when the user asks for a recap or a written summary.
Codebase-exploration shortcuts
Reach for these before asking the human:
| Question shape | Tool |
|---|---|
| "Where is X defined?" / "Who imports X?" | codemap query · Glob (file paths) |
| "What does file X export?" | codemap query --recipe find-symbol-definitions · Read the file |
| "What's the blast radius of X?" | codemap impact / query_recipe fan-in |
| "What conventions apply to this layer?" | Read the relevant rule under .agents/rules/ |
| "How does the existing similar module do this?" | codemap query for the pattern, then Read |
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 · 50 lines · 24 tokens per session scan A 7b29785f1579
grilling is a skill published in the GitHub repository stainless-code/codemap (8 stars, last pushed 6d ago), licensed MIT. It adds 24 tokens to every session and 816 once invoked, about $0.0001 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 skills, from other repositories
roam
Codebase comprehension via roam-code CLI. Use when exploring codebases, planning modifications, debugging failures, assessing PR risk, or checking architecture health. Triggers on: understanding project structure, pre-change safety checks, finding symbols/files, blast radius analysis, affected tests, health scoring…
analyze-usage
Use when asked to analyze codescout tool usage, check for error patterns, audit tool health, generate a usage report, spot anti-patterns, or clear usage statistics to start fresh.
m1nd-first
Use when investigating a repository, searching for implementation, reviewing changes, working from specs/docs, or preparing a risky code change in an environment where m1nd is available. This doctrine makes m1nd the first investigative layer before grep, glob, or manual file reads, except when the task is pure…
mcp-headless-test
Run live tests of the Ariadne MCP server using Claude Code in headless mode. Tests tool discovery, invocation, and filtering. Use when validating MCP server changes or checking integration with Claude Code.
mcp-analytics
Check Ariadne MCP tool usage analytics — session counts, tool call stats, durations, and error rates.
release
Prepare a new release of Ariadne packages. Creates changesets, verifies build/tests, and drafts the PR to main.