What is consult-mcp CLAUDE.md?

A project instruction file that tells Claude Code how the consult-mcp project is structured, written, tested, and planned.

Use it when developing or reviewing consult-mcp code, tests, personas, configuration files, or implementation plans.

It keeps coding work consistent with the project's agreed rules, such as asynchronous file and network work, logging style, and module size limits.

consult-mcp CLAUDE.md — Instructions file for Claude Code by oblogin: A project instruction file that tells Claude Code how…

Install

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.

agentmods

npx agentmods add instructions/oblogin/consult-mcp/claude-md

Clone the repo

git clone --depth 1 https://github.com/oblogin/consult-mcp

Made for: Claude Code.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/oblogin/consult-mcp/claude-md.svg)](https://agentmods.dev/instructions/oblogin/consult-mcp/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/oblogin/consult-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/oblogin/consult-mcp/claude-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 691 This file is loaded in full into every session.

When invoked 691 The same file — it is already loaded in full.

Security scan A 1 finding. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00691	$0.00691
Opus 5	$0.00345	$0.00345
Sonnet 5	$0.00138	$0.00138
Haiku 4.5	$0.00069	$0.00069

Measured 5d ago against content hash bc53d674290c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

consult-mcp CLAUDE.md scanned grade A with 1 finding 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.

Runs shell commandslowCapability

Expected in a hook, worth knowing in a rule or an instructions file.

- Не используй sync subprocess (`subprocess.run`, `os.system`) в production code paths. Только `asyncio.create_subprocess_exec`.

CLAUDE.md · 35 lines

How it starts

The opening of the file, as written. The whole thing — 35 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md — consult-mcp project conventions

Code style

Line length: 120 символов. Не 80, не 88.
Async I/O: любая работа с subprocess / файлами / сетью — через asyncio (asyncio.create_subprocess_exec, aiofiles если потребуется). Синхронный I/O допустим только в холодном bootstrap-коде (модульный импорт).
Persona format: markdown с YAML frontmatter (name, description, tags). Без каких-либо других форматов.
Logging: стандартный logging модуль Python. Без structlog/loguru в MVP. Логгер именуется logging.getLogger("consult_mcp.<submodule>").
LOC budget per module: ≤ 200 строк (без учёта docstrings и blanks). Если превышаем — разбиваем на подмодули.

Architecture notes

Layout: src/consult_mcp/ (PEP-621 src-layout).
Зависимости: только mcp>=1.0.0 в основном пакете. Dev-зависимости в [project.optional-dependencies].dev.
Entry-point: consult-mcp = "consult_mcp.server:main".
Персоны: built-in в src/consult_mcp/personas/*.md, user override в ~/.consult-mcp/personas/*.md.
CLI configs: built-in JSON в src/consult_mcp/conf/*.json, user override в ~/.consult-mcp/agents/*.json.

Testing

pytest + pytest-asyncio (asyncio_mode = "auto" в pyproject).
Smoke-тесты в tests/test_smoke.py могут зависеть от наличия CLI на машине — пропускать через pytest.skip при отсутствии.

Plans

Атомарные задачи и обзорные документы итераций — в Docs/plans/iterations/. Перед началом работы открой соответствующий *.md файл блока/задачи.

What NOT to do

Не добавляй HTTP-based LLM провайдеров — мы CLI-only. Это архитектурный выбор, не недоделка.
Не добавляй mutation/permission-bypass флаги (--permission-mode acceptEdits, --yolo, --dangerously-bypass-approvals-and-sandbox) в internal_args CLI конфигов в IT-001. Эти флаги переезжают в mutation_args в IT-002/C-01 и активируются ТОЛЬКО для тула implement. Это инвариант безопасности: consult и consult_parallel не могут случайно дать агенту разрешение писать файлы.
Не используй sync subprocess (subprocess.run, os.system) в production code paths. Только asyncio.create_subprocess_exec.
Не пиши multi-paragraph docstrings на tools — FastMCP использует docstring как описание для AI-клиента, длинные docstrings раздувают context.
Не выводи логи на stdout — он зарезервирован под MCP JSON-RPC frames. Только sys.stderr через logging.

Read the full file on GitHub · 35 lines

Changes

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 · 35 lines · 691 tokens per session scan A bc53d674290c

Subscribe to this mod's changes

consult-mcp CLAUDE.md is an instructions file published in the GitHub repository oblogin/consult-mcp (0 stars, last pushed 4mo ago), licensed MIT. It adds 691 tokens to every session, about $0.0035 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

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).

microsoft/vscode · 6,785 tokens

consult-mcp CLAUDE.md