Built for Claude Code, OpenCode, Cursor & Grok

Find What Slows Your
AI Coding Agent Down.

A Rust CLI & TUI that measures the hidden cost of every agent turn — subshell startup latency, MCP tool-schema tokens, instruction bloat, and skill collisions. Every number is measured, never guessed.

$ brew tap dautovri/tap && brew install agentprof
Subshell Tool Latency
617.6ms 5.3ms
⚡ Interactive vs non-interactive, measured here
50 Tool Calls Turnaround
+30.8s wait 0.26s
⚡ What 50 calls cost; recoverable with an opt-in guard
MCP Schema Load Per Turn
33,869 tok 12,153 tok
⚡ Measured; after disabling the 2 heaviest servers

The Complete Agent Optimization Stack

Written in Rust. Every figure it reports is measured, never guessed.

Subshell Latency Accelerator

Every AI agent tool execution forks a subshell. If your ~/.zshrc runs slow version managers (nvm, pyenv) or git status prompts, 50 tool executions waste over 30 seconds. agentprof injects an idempotent, zero-overhead bypass header so agent subshells run in under 6ms.

Subshell Init Benchmark 617ms ➔ 5.3ms (99.1% saved)
🔌

MCP Tool Schema Profiler

Connected Model Context Protocol servers inject extensive JSON tool definitions into your prompt on every turn. agentprof mcp profiles schema token overhead and flags heavyweight tools before they blow through your budget.

Schema Load ghidra: 14.1k tok • lldb: 3.4k tok
🎯

Skills & Trigger Collision Auditor

Audits all installed skills across Claude Code, OpenCode, and local worktrees. Identifies keyword overlaps where multiple skills compete for identical intents (e.g. review, qa, design) and cause hallucination loops.

Collision Guard 76 Skills • 6 Trigger Overlaps
🔍

Instruction Linter & Contradiction Detector

Detects conflicting rules across AGENTS.md, CLAUDE.md, and .cursorrules (e.g. ObservableObject vs @Observable, deployment target mismatches). Compresses bloated rule files and compiles them into context-routed JIT instructions.

Rule Conflict Linter Conflicts pinpointed by file & line
🛡️

Workspace Hygiene & Security Shield

Prevents AI agent search and grep tools from indexing massive unignored directories (.build/, DerivedData/, node_modules/, target/) or accidentally ingesting sensitive credential files (.env, *.pem, credentials.json). Merges patterns into .claudeignore and .cursorignore without discarding existing rules, keeping a backup.

1-Click Guard agentprof fix --dry-run (preview) → fix --ignore (merges, never overwrites)

Estimate Your Team's Agent Overhead

Estimate the overhead agentprof surfaces across a team, so you can decide what is worth cutting.

Daily Agent Tool Calls (per dev) 50
Active Developers in Team 5
Active MCP Servers 4
Monthly Developer Time Lost To Shell Startup
0.9 hrs
Monthly Token Cost Of MCP Schemas
$165.00
Assumes a 600ms subshell tax, 2,500 schema tokens per MCP server, and Claude Sonnet input pricing ($3/Mtok). Run agentprof bench and agentprof mcp --probe for your own numbers.