CtxWise

Stop burning 30,000+ startup tokens on idle MCP schemas, diverging CLAUDE.md / AGENTS.md files, and hidden prompt leaks.
The local-first AI Context Firewall, MCP Schema Tree-Shaker, and Token Profiler for Claude Code, OpenAI Codex, and Cursor.

Quickstart · GitHub Action · MCP X-Ray · Security Shield · Contribute a Preset · Benchmarks (-73.3% prompt)

Why Developers Install CtxWise

Modern coding agents (Claude Code, OpenAI Codex CLI, Cursor) silently load CLAUDE.md, AGENTS.md, .cursor/rules, skill catalogs, and MCP server tool schemas on every single turn:

1. Idle MCP Schema Bloat: Connecting 4–5 MCP servers (GitHub, Postgres, Playwright, Notion) can burn 15,000–40,000 tokens before you type a single word.
2. Multi-Agent Drift: Teams using both Claude Code (CLAUDE.md) and OpenAI Codex (AGENTS.md) end up with out-of-sync rules and duplicated boilerplate.
3. Silent Secret & Prompt-Injection Leaks: Hardcoded API keys (sk-ant-..., sk-proj-..., ghp_...) or covert HTML/Unicode instructions inside third-party skills get injected directly into model prompts.

CtxWise fixes all three locally in milliseconds — with zero model calls, zero API keys, and zero telemetry.

30-Second Quickstart

Run directly with npx in any repository where you use Claude Code, OpenAI Codex, or Cursor:

# 1. Audit startup tokens & parity across Claude Code (CLAUDE.md), Codex (AGENTS.md), and Cursor
npx @framy2/ctxwise agents

# 2. Scan context files & skills for leaked API keys, prompt injection, and hidden Unicode
npx @framy2/ctxwise shield --fail-on-risk

# 3. Profile MCP servers & find bloated tool schemas eating your context window
npx @framy2/ctxwise mcp-xray --probe-stdio

# 4. Sync CLAUDE.md <-> AGENTS.md <-> Cursor rules and apply lossless Token Diet compression
npx @framy2/ctxwise sync --compress --cursor

Key Capabilities

1. Multi-Agent Context Audit & "Token Diet" Sync (agents & sync)

Keep Claude Code (CLAUDE.md, .claude/skills, .claude/commands, .mcp.json), OpenAI Codex (AGENTS.md, .codex/skills), and Cursor (.cursor/rules/*.mdc) 100% aligned while stripping markdown bloat (comments, duplicate bullet rules, blank padding) without touching code blocks:

npx @framy2/ctxwise agents
npx @framy2/ctxwise sync --compress --cursor

2. MCP Schema Tree-Shaking (mcp-xray)

Discover all MCP servers declared across .mcp.json, ~/.claude.json, and ~/.codex/config.toml. Measure exact token weights per tool schema (either from a saved tools/list JSON file or by probing local stdio servers live) and spot [BLOATED] tools:

npx @framy2/ctxwise mcp-xray --probe-stdio

CtxWise mcp-xray · 3 MCP servers declared · 18 tool schemas analyzed · ~6,420 schema tokens (estimated)

Heaviest MCP tool schemas:
- ~890 tokens · github/create_or_update_pull_request (16 params, 3,560 chars) [BLOATED]
- ~640 tokens · postgres/execute_migration_plan (14 params, 2,560 chars) [BLOATED]

Tree-shaking potential: ~1,530 tokens/turn savable by pruning bloated tools.

3. Context Security Shield (shield)

Protect your repository before an agent reads your workspace or before a PR merges:

• Secret Exposure Detection: Catches hardcoded Anthropic (sk-ant-...), OpenAI (sk-proj-...), GitHub (ghp_...), AWS (AKIA...), and private key blocks inside CLAUDE.md, AGENTS.md, .mcp.json, and skills.
• Prompt Injection & Trojan Unicode Guard: Detects override directives (ignore all previous instructions), covert HTML comment payloads (<!-- curl ... -->), and hidden bidirectional/zero-width Unicode characters.

npx @framy2/ctxwise shield --fail-on-risk

4. Curated Stack Presets (preset)

Bootstrap a reviewable token budget (ctxwise.budget.yaml) and lean starter AGENTS.md + CLAUDE.md files tailored to your stack:

npx @framy2/ctxwise preset
npx @framy2/ctxwise preset nextjs-typescript --with-guidance

Built-in presets include:

• nextjs-typescript (Next.js + TypeScript + Tailwind)
• python-fastapi (Python 3.12 + FastAPI + Pydantic v2)
• rust-cargo (Rust + Cargo Workspace)
• go-microservice (Go 1.23+ Service)
• mcp-server-dev (Model Context Protocol Server)
• monorepo-turbo (Turborepo / pnpm Monorepo)

GitHub Action (CI Guard)

Add CtxWise to .github/workflows/ci.yml in 5 lines to automatically block secret leaks, prompt injection, and context token bloat on every Pull Request:

- name: AI Context Firewall & Token Budget Guard
  uses: FramY2/ctxwise@v1
  with:
    max-known-tokens: "4000"
    fail-on-secrets: "true"
    check-agent-parity: "true"

See what Codex carries in one command (snapshot & budget)

npx @framy2/ctxwise snapshot
npx @framy2/ctxwise budget --max-known-tokens 4000 --fail-on-exceed

Full CLI Reference

Contribute a Stack Preset in 5 Minutes

We welcome community contributions! Adding a new stack preset in presets/ (e.g., sveltekit, django, laravel, kotlin-spring, terraform-aws, flutter) takes less than 5 minutes and is a great Good First Issue / Hacktoberfest PR:

1. Add presets/<your-stack>.yaml following CONTRIBUTING.md.
2. Run npm test to validate your YAML schema automatically.
3. Open a Pull Request!

Live Evidence

The public microbenchmark ran 10 paired tasks across requested GPT-5.6 Luna, Terra, and Sol profiles: 20/20 answers passed the same exact validator. Removing irrelevant discovered context reduced the estimated model-visible prompt by 73.3% and exact aggregate turn tokens by 29.1% (with a second maintainer repeat measuring 28.5% across nine comparable pairs).

Read the method and limitations, inspect the machine-readable summary, or watch the 20-second MP4.

Cost Honesty by Design

See Architecture, Cost semantics, and Privacy and security.

Development

npm ci
npm run check
npm run build
npm run validate:plugin
npm pack --dry-run

License

Apache-2.0. See LICENSE and NOTICE.