verification-strategy — Claude Code skill: overview and setup
verification-strategy — a structured checklist skill for Claude Code to audit file outputs, compiler flags, exit codes, and edge cases before declaring a task finished.
MIT ★ 1,125 updated 2026-10-27
npx skills add vstorm-co/pydantic-deepagents --skill verification-strategy -a claude-codeWhat verification-strategy does
This skill forces Claude Code to run a thorough verification pass after writing code. The agent re-reads the exact prompt requirements, confirming specific compiler flags, absolute file paths, and output formats instead of assuming the work is done.
The process uses standard shell commands like wc -c, file, and echo $? to inspect byte counts, file encodings, and exit codes. If test scripts exist in the repository, the skill directs the model to read them first. This exposes hidden constraints such as strict delimiter rules or float precision.
Without a working shell environment to inspect stderr and compile binaries, the verification steps cannot run.
What it helps with
- verify exit codes and standard error before reporting completion
- confirm generated file sizes and paths against exact task requirements
- inspect test scripts to identify schema and precision constraints
- catch leftover debug logs and temporary files prior to finishing
Example prompts for the agent
- Apply verification-strategy before reporting that this task is complete
- Verify all generated outputs and compiler exit codes against the initial prompt
- Review the local test runner script and check edge cases using verification-strategy
How to install verification-strategy
1Skills CLI (Vercel)
npx skills add vstorm-co/pydantic-deepagents --skill verification-strategy -a claude-code2Manually: put the files in place
~/.claude/skills/verification-strategy/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/verification-strategy/
1Skills CLI (Vercel)
npx skills add vstorm-co/pydantic-deepagents --skill verification-strategy -a codex2Manually: put the files in place
.agents/skills/verification-strategy/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add vstorm-co/pydantic-deepagents --skill verification-strategy -a cursor2Manually: put the files in place
.cursor/skills/verification-strategy/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install vstorm-co/pydantic-deepagents apps/cli/skills/verification-strategy2Manually: put the files in place
.github/skills/verification-strategy/SKILL.md1Skills CLI (Vercel)
npx skills add vstorm-co/pydantic-deepagents --skill verification-strategy -a gemini-cli2Platform CLI
gemini skills install https://github.com/vstorm-co/pydantic-deepagents --path apps/cli/skills/verification-strategyManually: put the files in place
.devin/skills/verification-strategy/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about verification-strategy
How to install verification-strategy in Claude Code?
Run npx skills add vstorm-co/pydantic-deepagents --skill verification-strategy -a claude-code or save the skill directly to ~/.claude/skills/verification-strategy/SKILL.md.
What does the verification-strategy skill do?
It provides a systematic checklist covering task instructions, output file checks, compilation commands, and edge cases before an agent finishes.
What common issues does the checklist prevent?
It catches discrepancies like printing to stdout instead of a target file, trailing whitespace differences, wrong decimal precision, and non-zero exit codes.
Source text
SKILL.md · MIT
---
name: verification-strategy
description: "Thorough verification of completed work before declaring done"
tags: [testing, verification, benchmark]
version: "1.0.0"
---
# Verification Strategy
How to verify your work is correct before declaring a task complete.
## The Verification Checklist
After implementing a solution, go through ALL of these:
### 1. Re-read the original task
- Re-read the EXACT wording of the task instruction
- Check every requirement — file paths, names, formats, constraints
- Are there implicit requirements? ("compile with gcc -O3 -lm" means exact flags)
### 2. Check file outputs
- Do all required files exist at the exact paths specified?
- Are file sizes within any stated limits?
- Is the file format correct? (binary vs text, encoding, line endings)
```bash
ls -la /path/to/expected/output
wc -c /path/to/output # byte count
file /path/to/output # format detection
head -5 /path/to/output # content preview
```
### 3. Compile and run
- Compile with the EXACT flags specified in the task
- Run with the EXACT command and arguments specified
- Check exit code: `echo $?` (should be 0 for success)
- Check both stdout AND stderr
### 4. Validate output
- Compare output against expected format
- Check exact field names, delimiters, number formatting
- If the task specifies output format, match it exactly:
- JSON: valid JSON? correct schema?
- CSV: correct headers? correct delimiter?
- Plain text: correct line endings? trailing newline?
### 5. Test edge cases
- Empty input (if applicable)
- The specific test inputs mentioned in the task
- Large inputs (if the task involves performance)
### 6. Check constraints
- Size limits (file size, code length)
- Time limits (does it finish in reasonable time?)
- Memory limits (does it stay within bounds?)
- No external dependencies that aren't available
## Reading Test Scripts
If you can find the test script, READ IT:
- What exact assertions does it make?
- What inputs does it use?
- What output format does it expect?
- What timeouts are set?
Tests often check things you didn't expect:
- Exact string matching (whitespace matters!)
- Specific numeric precision
- File permissions
- Process exit codes
## Common Verification Failures
| What you think is right | What the test actually checks |
|------------------------|-------------------------------|
| Output on stdout | Output in a specific file |
| Human-readable numbers | Exact decimal precision |
| Relative file paths | Absolute file paths |
| "Close enough" answer | Exact match |
| Code that runs | Code under a specific size limit |
## Final Check
Before declaring done, ask yourself:
1. Did I create/modify ALL the files the task asks for?
2. Did I use the EXACT paths, names, and formats specified?
3. Does my code compile/run WITHOUT errors?
4. Does my output match what automated tests would expect?
5. Have I cleaned up any debug output or temporary files?
© vstorm-co · MIT · vstorm-co/pydantic-deepagents