How do I install the Verification skill?

Run npx skillmds@latest add rom-orlovich/verification-2 in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Verification safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Verification?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Verification free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Verification?

rom-orlovich (@rom-orlovich) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Verification by rom-orlovich — Agent Skill

Name: Verification
Rating: 4.2 (5 reviews)
Author: rom-orlovich

File contents

Verification Skill

Run scripts → Score objectively → No opinions.

Quick Reference

Templates: See templates.md for reporting verification results

Core Rule

NEVER approve without running scripts. Exit codes are truth.

Scripts (Mandatory)

# Run ALL before scoring
.claude/scripts/verification/test.sh      # 40% weight
.claude/scripts/verification/build.sh     # 20% weight
.claude/scripts/verification/lint.sh      # 20% weight
.claude/scripts/verification/typecheck.sh # 20% weight

Score Calculation

Score = (test×0.4) + (build×0.2) + (lint×0.2) + (typecheck×0.2)
Where: pass=100, fail=0

Total Score	Decision
≥90%	APPROVE
<90%	REJECT with gaps

Verification Steps

Read PLAN.md criteria
Run each script, capture exit code + output
Map criteria to script results
Calculate weighted score
Format response (APPROVE or REJECT)

Script Result Recording

Script	Exit	Weight	Score
test.sh	0/1	40%	0/40
build.sh	0/1	20%	0/20
lint.sh	0/1	20%	0/20
typecheck.sh	0/1	20%	0/20
Total	-	100%	X%

Gap Analysis (For Rejections)

Each gap must include:

Gap: [Title]
- Script: Which failed
- Output: Relevant error
- Agent: Who should fix
- Fix: Specific instruction

Iteration Handling

Iteration	Tone
1	Detailed, educational
2	Direct, focused on remaining
3 (final)	Force decision

At iteration 3: Either "Deliver with caveats" OR "Escalate to user"

Anti-Patterns

Don't	Do
"Looks good to me"	Run scripts, show exit codes
Approve to end loop	Score based on evidence
Vague feedback	Specific file:line fixes

---name: verification-2description: Script-based verification with confidence scoring. Use when verifying code quality, running linting/type checking, or validating code changes before completion.---# Verification Skill> Run scripts → Score objectively → No opinions.## Quick Reference- **Templates**: See [templates.md](templates.md) for reporting verification results## Core Rule**NEVER approve without running scripts.** Exit codes are truth.---## Scripts (Mandatory)```bash# Run ALL before scoring.claude/scripts/verification/test.sh # 40% weight.claude/scripts/verification/build.sh # 20% weight.claude/scripts/verification/lint.sh # 20% weight.claude/scripts/verification/typecheck.sh # 20% weight```## Score Calculation```Score = (test×0.4) + (build×0.2) + (lint×0.2) + (typecheck×0.2)Where: pass=100, fail=0```| Total Score | Decision || ----------- | ---------------- || ≥90% | APPROVE || <90% | REJECT with gaps |---## Verification Steps1. **Read** PLAN.md criteria2. **Run** each script, capture exit code + output3. **Map** criteria to script results4. **Calculate** weighted score5. **Format** response (APPROVE or REJECT)---## Script Result Recording| Script | Exit | Weight | Score || ------------ | ---- | ------ | ------ || test.sh | 0/1 | 40% | 0/40 || build.sh | 0/1 | 20% | 0/20 || lint.sh | 0/1 | 20% | 0/20 || typecheck.sh | 0/1 | 20% | 0/20 || **Total** | - | 100% | **X%** |---## Gap Analysis (For Rejections)Each gap must include:```Gap: [Title]- Script: Which failed- Output: Relevant error- Agent: Who should fix- Fix: Specific instruction```---## Iteration Handling| Iteration | Tone || --------- | ---------------------------- || 1 | Detailed, educational || 2 | Direct, focused on remaining || 3 (final) | Force decision |**At iteration 3:** Either "Deliver with caveats" OR "Escalate to user"---## Anti-Patterns| Don't | Do || ------------------- | ---------------------------- || "Looks good to me" | Run scripts, show exit codes || Approve to end loop | Score based on evidence || Vague feedback | Specific file:line fixes |

Frequently asked questions

Verification