What is code-impact-mcp CLAUDE.md?

Instructions for a TypeScript MCP server that maps imports and exports in a JavaScript or TypeScript project to show which files may be affected by a change. A dependency graph records those relationships.

Use them when building, testing, linting, smoke-testing, releasing, or changing its dependency-impact and cycle-detection analysis.

They provide the commands, architecture, release files, and design limits needed to develop and publish the tool correctly.

code-impact-mcp CLAUDE.md — Instructions file by vk0dev

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/vk0dev/code-impact-mcp/claude-md

Clone the repo

git clone --depth 1 https://github.com/vk0dev/code-impact-mcp

Per session 390 This file is loaded in full into every session.

When invoked 390 The same file — it is already loaded in full.

Security scan A 0 findings. 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	$0.00390	$0.00390
Opus 5	$0.00195	$0.00195
Sonnet 5	$0.00078	$0.00078
Haiku 4.5	$0.00039	$0.00039

Measured 2d ago against content hash 67eba85665a4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

code-impact-mcp CLAUDE.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 2d 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.

CLAUDE.md · 33 lines

What it actually says

@vk0/code-impact-mcp

Dependency graph and blast-radius analysis for agent-driven code changes. Predicts what breaks before you commit.

Stack

TypeScript, @modelcontextprotocol/sdk, Zod, ts-morph
Vitest for unit tests

Commands

Install: npm ci
Build: npm run build
Test: npm test
Lint: npm run lint
Smoke test: npm run smoke

Publishing

See PUBLISHING.md for full marketplace playbook
Version sync: package.json, .claude-plugin/plugin.json, server.json, src/createServer.ts
Release: git tag -a vX.Y.Z && git push --follow-tags → CI publishes to npm + MCP Registry + Smithery

Architecture

src/createServer.ts — McpServer factory, exports createServer() and createSandboxServer() (Smithery)
src/server.ts — CLI/stdio entry point, uses createServer()
src/index.ts — Barrel exports for library usage
src/tools/index.ts — 4 MCP tools registration
src/graph.ts — ts-morph dependency graph builder, impact analysis, cycle detection

Notes

Analyzes TS/JS projects by parsing import/export relationships with ts-morph
Graph is cached per (projectRoot, tsconfigPath) pair
Local-first, zero network egress
Verification note: npm test exits cleanly on a clean/stashed tree. If tracked runtime/test files like src/graph.ts or tests/graph.test.ts are left dirty from another lane, isolate them with git stash before QA so you do not misread a dirty-worktree timeout/regression as a repo-wide vitest exit bug.

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.

2d ago First seen · 33 lines · 390 tokens per session scan A 67eba85665a4

Subscribe to this mod's changes

code-impact-mcp CLAUDE.md is an instructions file published in the GitHub repository vk0dev/code-impact-mcp (1 stars, last pushed 8d ago), licensed MIT. It adds 390 tokens to every session, about $0.0019 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.

Debugging ai-agent anthropic blast-radius claude-code mcp mcp-server

agent-bom CLAUDE.md

Instructions for msaad00/agent-bom, covering claude-specific addendum, memory, tools and communication.

msaad00/agent-bom · 411 tokens

coderadius AGENTS.md

Instructions for coderadius-ai/coderadius, covering agents.md, what this is, build & run commands, testing and architecture.

coderadius-ai/coderadius · 5,788 tokens

menhir AGENTS.md

AGENTS.md instructions for Archolith/menhir, covering menhir agent instructions, start here, implementation rules and verification.

Archolith/menhir · 395 tokens

Spiderbrain.ai-Coremind AGENTS.md

Instructions for aabhisrv/Spiderbrain.ai-Coremind: This repo carries a committed brain in .spiderbrain/: a deterministic, source-free map of its structure, dependencies, and blast radius (36 files, 20 edges). Consult it before reading files, to know what matters and what a change reaches.

aabhisrv/Spiderbrain.ai-Coremind · 443 tokens

menhir CLAUDE.md

Claude Code instructions for Archolith/menhir: Follow AGENTS.md. Start with .agent/README.md and use its task router; do not read the entire .agent/ tree by default.

Archolith/menhir · 92 tokens

menhir copilot-instructions.md

Copilot instructions for Archolith/menhir: Follow AGENTS.md. Start with .agent/README.md and use its task router rather than reading all project documentation. Prefer structural queries before filesystem search, check blast radius before edits, preserve unrelated worktree changes, and suggest focused affected tests.

Archolith/menhir · 61 tokens