Borrowing it
Nothing to install: this file belongs to getmasv/masv-mcp-server. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/getmasv/masv-mcp-server/main/.kiro/steering/tool-design.mdgit clone --depth 1 https://github.com/getmasv/masv-mcp-serverWrote 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/getmasv/masv-mcp-server/tool-design)<a href="https://agentmods.dev/instructions/getmasv/masv-mcp-server/tool-design"><img src="https://agentmods.dev/badge/instructions/getmasv/masv-mcp-server/tool-design/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/instructions/getmasv/masv-mcp-server/tool-design"><img src="https://agentmods.dev/badge/instructions/getmasv/masv-mcp-server/tool-design.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 | $0.00814 | $0.00814 |
| Opus 5.5 | $0.00326 | $0.00326 |
| Sonnet 5.5 | $0.00163 | $0.00163 |
| Haiku 4.5 | $0.00081 | $0.00081 |
Grade A, and why
masv-mcp-server tool-design.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 26d 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Tool Design
The highest-leverage surface in this project. Models pick tools with well-written descriptions far more often, so a capability the manifest does not clearly state effectively does not exist. A description defect is indistinguishable from a missing feature — treat naming and wording as functional work, not polish.
Name for the user's job, not the API shape
A tool name is read by a model deciding what to call. transfer_files_from_integration does not match
"send this folder to my colorist."
| Avoid | Prefer |
|---|---|
transfer_files_from_integration |
send_files |
list_files_on_integration |
browse_files |
send_package_to_integration |
deliver_package_to_storage |
get_integrations |
list_connected_sources |
Descriptions carry the user's vocabulary
For the send path the description must contain: send, share, download link, folder, people, email, my computer, NAS, Desktop, no file size limit. Name Storage Gateway and illustrate it — cloud-only examples teach the model that local disk is out of scope.
- Lead with the outcome ("send files to people"), not the mechanism ("to MASV").
- State the prerequisite up front (MASV Desktop or Agent connected as a gateway).
- State limits and non-limits explicitly.
- Surface anything the model cannot infer. A provider string like
desktop_sgmeans nothing to a model that was never told what it is.
Annotations are mandatory
Every tool gets a title, an explicit readOnlyHint, an explicit destructiveHint whenever
readOnlyHint is false, and idempotentHint where genuinely true.
Not cosmetic: required for the Claude directory, required for Microsoft, a named top rejection cause on OpenAI, and unannotated tools trigger an approval prompt on every single call in Cowork.
Two traps:
destructiveHintdefaults totrue. Omitting it on a write tool silently marks it destructive and invites prompts on routine actions. Always set it explicitly.- Annotate honestly, by effect on state.
destructiveHintmeans "may mutate or delete durable state". Do not use it as a generic danger flag for expensive or irreversible-in-other-ways operations; see the annotation table inarchitecture.md. Dishonest labels get audited, and they degrade unattended use.
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.
- 26d ago First seen · 70 lines · 814 tokens per session scan A a94daf7d7f08
masv-mcp-server tool-design.md is an instructions file published in the GitHub repository getmasv/masv-mcp-server (5 stars, last pushed 21d ago), licensed MIT. It adds 814 tokens to every session, about $0.0033 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 2026-09-25.
Other instructions, from other repositories
vibetags com-example-config-AsyncTestConfig.instructions.md
Instructions for PIsberg/vibetags, covering copilot instructions for asynctestconfig and immutable type.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).