Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

Command Optimization Workflow · nwave-ai | DecimalAI

agent toolingby nwave-ai

Command Optimization Workflow

nwave-ai/nw-command-optimization-workflow

Source nwave-ai/nwave ↗License MIT

Step-by-step workflow for converting bloated command files to lean declarative definitions

SkillScore87Maintained

On-demandOn-demand: invoked explicitly — never occupies your agent's context until called.

AI Engineer pack

Last verified 8/16/2026

What it changes

Passed 19 of 22 test cases with the skill, and 6 of 22 without it.

Without skill26%

With skill87%

Checks it passed that the model alone missed

The response caps the example count at 3 max.
The analysis assigns a command classification.
The summary report includes before and after line counts.

and 15 more in the full report

15 fixed5 passed either way2 still failing1 broke

Pass-rate lift: +61 pp
Tokens: +24%
Agent turns: 0%
Response time: -43%

Model gemini-3.6-flashMeasured 8/16/2026v120 of 23 cases passed with skill · 22 comparable

Tokens, turns and time describe resource use, not a dollar price.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/nwave-ai-nw-command-optimization-workflow/SKILL.md https://app.decimal.ai/s/nwave-ai-nw-command-optimization-workflow/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add nwave-ai-nw-command-optimization-workflow, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/nwave-ai-nw-command-optimization-workflow/ and pins it in .decimal/skills.lock

decimalai skills add nwave-ai-nw-command-optimization-workflow

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/nwave-ai-nw-command-optimization-workflow/SKILL.md https://app.decimal.ai/s/nwave-ai-nw-command-optimization-workflow/SKILL.md

Raw file: /s/nwave-ai-nw-command-optimization-workflow/SKILL.md ↗ · machine-readable summary at /s/nwave-ai-nw-command-optimization-workflow.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 14 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 14

Prompt

The file `develop.md` is currently over 2,300 lines long. When refactored to pure high-level orchestration, what is the line count target?

What was checked, without and with the skill
Checked	Without skill	With skill
The response sets the target line count for develop.md between 200 and 300 lines.	Missed	Passed

case-19 · gemini-3.6-flash · 8/16/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md91 linessingle file

Command Optimization Workflow

Overview

Convert oversized commands (500-2400 lines) to lean declarative definitions (40-300 lines) by removing duplication, extracting domain knowledge to agents, and adopting the forge.md pattern.

Phase 1: Analyze

Classify: dispatcher, orchestrator, or simple
Measure: wc -l the file
Identify content categories (approximate %): parameter parsing/validation | Workflow/orchestration logic | Agent prompt templates | Examples/documentation | Error handling | Boilerplate (orchestrator briefing, agent registry)
Flag reducible content: Duplication from other commands (orchestrator briefing, agent registry, parameter parsing) | Domain knowledge belonging in agents (TDD phases, review criteria, methodology) | Dead code (deprecated formats, aspirational features, old signatures) | Verbose examples (JSON blocks exceeding 3 examples) | Aggressive language (count CRITICAL/MANDATORY/MUST)

Phase 2: Extract

2a: Remove shared boilerplate

These blocks appear in 5-12 commands, extract to shared orchestrator preamble skill:

"Sub-agents have NO ACCESS to Skill tool" briefing (~20-30 lines/file)
Agent registry with capabilities (~15-20 lines/file)
Parameter parsing rules (~10-15 lines/file)
Pre-invocation validation checklist (~10-15 lines/file)
"What NOT to include" blocks (~8-12 lines/file)

After extraction, command references preamble.

2b: Move domain knowledge to agents

If command contains HOW the agent works, move to agent definition or skill:

TDD phase details -> nw-software-crafter | DIVIO templates -> nw-documentarist | Refactoring hierarchies -> nw-software-crafter | BDD/Gherkin -> nw-acceptance-designer | Review criteria -> reviewer agent | Mutation testing config -> nw-software-crafter

Command retains WHAT and success criteria. Agent owns HOW.

2c: Delete dead content

Remove: deprecated format references | Aspirational unimplemented features | Verbose JSON contradicting current format | BUILD:INJECT placeholders | Mixed-language comments

2d: Reduce aggressive language

"CRITICAL: You MUST extract..." -> "Extract..."
"MANDATORY: Use the Task tool" -> "Use the Task tool"
"NON-NEGOTIABLE requirement" -> Remove adjective entirely

2e: Condense examples

Track record

Benchmark

19 / 22

test cases passed, vs 6 / 22 without

Adoption

Users: 0
Times used: 0

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

nwave-ai / nwave

nWave/skills/nw-command-optimization-workflow/SKILL.md

View on GitHub ↗

License MITImported 8/3/2026Version v1

More skills from the same source and nearby in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

See it in action

What's inside

Command Optimization Workflow

Overview

Phase 1: Analyze

Phase 2: Extract

2a: Remove shared boilerplate

2b: Move domain knowledge to agents

2c: Delete dead content

2d: Reduce aggressive language

2e: Condense examples

Track record

Benchmark

Adoption

SkillSafety

Source

Related skills