Stop agents from faking progress — demand proof in every status update
Agents love to say 'on it!' or 'done!' when nothing has actually run. Bind every status update to a process ID, file path, URL, or command output.
- Type: Prompt
- Tags: Coding Agent Hacks, Intermediate, Claude Code, Cursor
- Updated: 2026-10-01
Prompt
Agents love to say 'on it!' or 'done!' when nothing has actually run. Bind every status update to a process ID, file path, URL, or command output. By default, agents give confident status updates whether the action worked or not — you won't catch silent failures until you check by hand. Tech founder Cathryn Lavery shared a one-line rule that fixes it: every claim must come with proof. Paste it into `CLAUDE.md` (project) or `~/.claude/CLAUDE.md` (global) for Claude Code, or `.cursorrules` for Cursor. Now 'building' becomes 'started build, pid 41822, branch feat/auth, step: tsc.' Implementation example: # CLAUDE.md / ~/.claude/CLAUDE.md / .cursorrules Never say 'done' or 'working on it' unless the action has actually started. Every status update must include proof — a process ID, file path, URL, or command output. No proof = didn't happen. A false completion is worse than a delayed honest answer.
More prompts
- Expense Report Automation System — Save More Time
- Meeting Scheduling Link Standardization — Save More Time
- Invoice Generation and Delivery Automation — Save More Time
- Document Template Automation Library — Save More Time
- Calendar Management Automation Rules — Save More Time
- File Organization and Naming System — Save More Time
- Recurring Payment Automation Setup — Save More Time
- Data Entry Elimination Framework — Save More Time