Exceeds Expectations Skills Marketplace
- 1 repo stars
- Updated March 13, 2026
- kevinshen56714/exceeds-expectations
exceeds-expectations is a Claude Code skill pack that escalates weak agent execution into evidence-first investigation, verification, and ownership before work is marked complete.
About
exceeds-expectations is a Claude Code skill pack that corrects agent failure modes like repeated guessing, weak verification, passive ownership, and premature done declarations. It activates when work has failed twice, responses drift toward excuses, or a surface fix ships without proof, and it organizes blunt operating standards around evidence, investigation, and closed-loop verification. Invoke /exceeds-expectations or let it trigger when quality slips. Use it when you want performance-management-style rigor from your coding agent before you accept a fix as complete.
- Dryly cynical coaching for evidence-first engineering behavior
- Triggers on repeated failures, hand-waving, or premature completion
- Six operating standards from execution quality to product judgment
- Recovery method: stop vibe-debugging, open receipts, close the loop
- Available for Claude Code, Codex CLI, and Cursor
Exceeds Expectations Skills by the numbers
- Data as of Aug 10, 2026 (Skillselion catalog sync)
/plugin marketplace add kevinshen56714/exceeds-expectationsAdd your badge
Show developers this marketplace is listed on Skillselion. Paste this into your README.
| repo stars | ★ 1 |
|---|---|
| Last updated | March 13, 2026 |
| Repository | kevinshen56714/exceeds-expectations ↗ |
What do I do when my coding agent keeps looping, hand-waving, or calling a patch done without proving the system actually works?
Escalate looping or hand-wavy agent work into evidence-first investigation, verification, and ownership before calling tasks done.
Who is it for?
Developers who want blunt execution coaching that forces receipts, verification, and ownership when agent quality slips.
Skip if: Casual exploratory chats where strict verification ladders and escalation standards would slow harmless experiments.
What you get
The agent follows the recovery method and expectations ladder, producing verified fixes with documented evidence instead of unchecked patches.
Plugins in this marketplace
1 plugin - install individually after you add the marketplace.
Recommended Marketplaces
Beast Modeaaronstatic is a Claude Code marketplace for testing.6
Ai Skill Test Bddai-skill-test-bdd is a Claude Code marketplace for testing.
Ai Test Failure AnalyzerAI Test Failure Analyzer is a Claude Code plugin that diagnoses failing tests by gathering evidence from git history, logs, and config rather than guessing at fixes.
Android Test Pilotandroid-test-pilot is a Claude Code plugin marketplace from panicgit/android-test-pilot that automates Android app testing with a text-first strategy cheaper and faster than screenshot-per-step mobile-mcp loops.1
Anti Hallanti-hall is a Claude Code plugin marketplace centered on a single verify-first discipline plugin for developers who ship with agents daily.1
Askqa Marketplaceaskqa-marketplace is a Claude Code marketplace for testing.FAQ
When should exceeds-expectations activate?
After repeated failures, unsupported excuses, user-delegated diagnostics, or a surface fix without verification—or when you invoke /exceeds-expectations.
What behaviors does it enforce?
Evidence before confidence, investigation before handoff, verification before completion, and ownership beyond the first fix.
What is the recovery method?
Stop vibe-debugging, open the receipts, do boring checks, kill your favorite theory, and close the loop.