I stopped building AI tools and started doing the tasks manually instead

Stephen Lee — Fri, 03 Apr 2026 05:02:06 +0000

I've been building AI products for a while now. Every time I thought I knew what to automate, I was wrong.

So I tried something different. Instead of building another tool, I put up a simple page: describe a repetitive task, pay $5, I do it and email you the result.

The tech stack is intentionally boring. Django backend, Stripe checkout, SQLite, static HTML frontend. No React. No AI. No accounts. The whole thing is under 500 lines of code.

Why? Because the product isn't the software. The product is the dataset of tasks people submit. Every submission gets categorized: input type, output type, difficulty, whether AI failed at it, whether it's recurring.

After enough submissions, patterns emerge. Those patterns become the first automated tools. The manual execution phase is just product discovery with real revenue.

The site is euphorie.com if you want to see how minimal it is.

Has anyone else tried using manual execution as a product discovery method? Curious what other approaches people have used to figure out what to build.

How I built a trust scoring hook for Claude Code

Stephen Lee — Sun, 08 Mar 2026 08:03:30 +0000

I've been using Claude Code for a while and realized I had zero visibility into what the agent was doing across sessions. Which tools it called, whether it touched files it shouldn't have, how many calls it made per task.

So I built a hook that scores every session.

What it does

The hook listens to three Claude Code events:

PostToolUse — records every tool call, checks it against an allowlist, flags protected path access
PreToolUse — blocks tool calls that touch sensitive files like .env or SSH keys
Stop — computes the final trust score and logs it

At the end of each session you get:

[authe.me] Trust Score: 92 (reliability=100 | scope=75 | cost=100)
[authe.me] tools=14 violations=1 failed=0

The scoring

Three dimensions, weighted into an overall score:

Reliability (40%) — what percentage of tool calls succeeded
Scope (35%) — did the agent stay within your allowed tools and paths. Each violation drops the score by 25 points
Cost (25%) — how many tool calls were made. Under 20 is fine, over 100 starts flagging

Hash chaining

Every tool call event gets hashed with the previous hash, creating a chain. If someone tampers with the log, the chain breaks. Same concept as blockchain audit trails but much simpler.

def compute_hash(prev_hash, data):
    payload = f"{prev_hash}:{json.dumps(data, sort_keys=True)}"
    return hashlib.sha256(payload.encode()).hexdigest()[:16]

Protected path blocking

The PreToolUse hook checks if Claude is about to read or edit a sensitive file. If it matches, the hook returns a deny decision with a reason that Claude receives as feedback:

result = {
    "hookSpecificOutput": {
        "hookEventName": "PreToolUse",
        "permissionDecision": "deny",
        "permissionDecisionReason": "authe.me policy: .env is a protected path."
    }
}

Claude sees the denial reason and adjusts its plan.

Install

mkdir -p ~/.claude/hooks ~/.authe
curl -fsSL https://raw.githubusercontent.com/autheme/claude-code-hook/main/authe-hook.py -o ~/.claude/hooks/authe-hook.py
chmod +x ~/.claude/hooks/authe-hook.py

Then add the hook config to ~/.claude/settings.json:

{
  "hooks": {
    "PostToolUse": [
      {
        "matcher": "",
        "hooks": [
          {
            "type": "command",
            "command": "AUTHE_HOOK_EVENT=PostToolUse python3 ~/.claude/hooks/authe-hook.py"
          }
        ]
      }
    ],
    "Stop": [
      {
        "matcher": "",
        "hooks": [
          {
            "type": "command",
            "command": "AUTHE_HOOK_EVENT=Stop python3 ~/.claude/hooks/authe-hook.py"
          }
        ]
      }
    ]
  }
}

Single file, zero dependencies, pure Python.

What's next

Working on remote reporting so you can aggregate scores across sessions and agents. Also building an OpenClaw plugin that does the same thing for that ecosystem.

Repo: github.com/autheme/claude-code-hook

Built with Claude. Would love feedback from anyone running Claude Code in production.

Forem: Stephen Lee