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 instructions/lykhoyda/ask-llm/agents-mdgit clone --depth 1 https://github.com/Lykhoyda/ask-llmWrote 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/instructions/lykhoyda/ask-llm/agents-md)<a href="https://agentmods.dev/instructions/lykhoyda/ask-llm/agents-md"><img src="https://agentmods.dev/badge/instructions/lykhoyda/ask-llm/agents-md.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 | $0.01120 | $0.01120 |
| Opus 5 | $0.00560 | $0.00560 |
| Sonnet 5 | $0.00224 | $0.00224 |
| Haiku 4.5 | $0.00112 | $0.00112 |
Grade A, and why
ask-llm AGENTS.md 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 today.
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 — 22 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Project agent memory
This file is the project's committed home for project-intrinsic agent knowledge: build, test, release, architecture, and sharp-edge notes that should travel with the code.
- Build/test:
yarn install --immutable, thenyarn build(topological — run it BEFOREyarn test; workspace tests resolve sibling packages fromdist/, so unbuilt@ask-llm/sharedfails every suite). Lint viayarn lint(biome format is enforced;biome check --writefixes). Harness-facing changes use the canonical local-onlyyarn prepr:harnessgate; opt-in live authorization and evidence rules are indocs/HARNESS-SMOKE.md. - Architecture decisions live in
docs/DECISIONS.md(numbered ADRs, newest prepended at the top); provider behavior differences are deliberate and cataloged indocs/PROVIDER-PARITY.md— read both before "aligning" providers. - Docs-model drift is CI-enforced:
scripts/check-docs-drift.mjsrequiresapps/docs/.vitepress/theme/providers.tsdefaultModel(and gemini/antigravityfallbackModel) values to match each package'sconstants.tsliterally, matched inside each provider's own object block. Change both together. - MCP Registry rejects duplicate versions. Keep normal and recovery publication routed through
scripts/publish-mcp-registry.mjs; it verifies exact existing records and publishes only missing ones (ADR-139). A manual registry retry must not re-enter npm publication. - Public package source tags are remote
<name>@<version>refs managed only byscripts/create-or-verify-package-tags.mjson normal and recovery release paths (ADR-151/152);changesets/actionv2 must keep bothcreate-github-releases: falseandpush-git-tags: false, while the helper excludes private shared, never retags/forces, and coexists with the single unifiedv*GitHub Release. - Gemini quota fallback is
gemini-3.7-flash(ADR-154); the antigravity fallback staysgemini-3.5-flashdeliberately — agy slugs are evidence-pinned to its live catalog (ADR-137), so never bump both in sympathy. - Provider default-model constants leak further than the provider package:
packages/llm-mcp/src/constants.ts(PROVIDERS.*.defaultModel) is threaded into executors asoptions.modelbymachine.ts, so executors cannot use "was a model passed?" to detect user pins (see ADR-137). - agy (Antigravity) CLI ground truth for this repo's executor is recorded in ADR-137 (minimum 1.1.5, base slug + separate
--effort, per-slug effort tiers, the live-capturedinvalid model selectionerror grammar) and ADR-141 (answers come from--output-format jsonstdout — keyresponse, usage object — on every supported version;--disable-slash-commandsis 1.1.9+ only). - Grok transport and model identity are separate (ADR-146/147;
/brainstormexact Grok + Sol routing and partial-consensus rules are in ADR-148): unified startup probes bothxai-apiandgrok-clireadiness because request-level harness selection happens later (an explicitASK_GROK_HARNESSnarrows readiness to that default), but execution selects exactly one explicit/default harness with no failover; Cursor Agent is a model-neutral harness (ask-cursor-agent) requiring separate provider + exact Cursor catalog ID, never a provider itself. @ask-llm/pluginis the one publishable Claude Code + Cursor Agent + Pi host package (ADR-142/147); host contracts live in the skill adapters,apps/docs/plugin/cursor.md, anddocs/PI-COMPATIBILITY.md. Cursor uses Agent Skills + a unified-onlymcp.json(oneask-llmentry; split Codex/Grok servers are optional user installs whose leaves are preferred when exposed) exposing exactlycodex-pair/grok-pair(explicit emptyagents/commands/hooksso Claude agents andhooks/hooks.jsonare not auto-discovered; other skills deferred), never Claude hooks/namespaces; the Claude plugin.mcp.jsonbundles only Codex while@ask-llm/mcp/@ask-llm/grok-mcpstay user-scoped, and pair skills call unifiedask-llmonly fully pinned; Pi excludesfable-review/grok-pair, uses native provider tools, and codex-pair requires marker + project trust + a user-owned canonical-project allowlist. The published Pi host floor must equal the oldest exactpi-package-smokematrix pin (currently 0.83.0).
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.
- today Changed 7b084b3f2f9d
- 4d ago First seen · 22 lines · 1,120 tokens per session scan A f6288ed35867
ask-llm AGENTS.md is an instructions file published in the GitHub repository Lykhoyda/ask-llm (16 stars, last pushed yesterday), licensed MIT. It adds 1,120 tokens to every session, about $0.0056 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 2026-08-30.
Other instructions, from other repositories
embedded-systems-architect AGENTS.md
Instructions for Zachariah9420/embedded-systems-architect: This repository ships an agent skill for embedded-systems and edge-AI development work.
qa-orchestra CLAUDE.md
Claude Code instructions for Anasss/qa-orchestra, covering qa orchestra — claude code instructions, what this project is, quick start, 1. clone into your workspace and 2. fill in your project context.
claude-sdlc CLAUDE.md
Claude Code instructions for lantisprime/claude-sdlc, covering claude.md, what this repo is, design intent — read this before "improving" anything, eat your own dog food and hook strictness philosophy.
claude-account-switch AGENTS.md
AGENTS.md instructions for LightZirconite/claude-account-switch, covering claude + codex account switch - agent guide, product invariants, authentication ux, switching safety and provider parity and deliberate differences.
Cortex CLAUDE.md
Instructions for Remskill/Cortex, covering claude.md, project overview, development commands, setup & database and docker.
aesth CLAUDE.md
Claude Code instructions for zircote-plugins/aesth, covering claude.md, architecture, key concepts, memory-first design and memory types.