skill-conductor

name: skill-conductor description: > Create, edit, evaluate, and package agent skills. Use when building a new skill from scratch, improving an existing skill, running evals to test a skill, benchmarking skill performance, optimizing a skill's description for better triggering, reviewing third-party skills for quality, or packaging skills for distribution. Not for using skills or general coding tasks.

Skill Conductor

Full lifecycle management for agent skills: draft → test → review → improve → repeat.

One skill to rule them all — from architecture to packaging. The core loop is always the same: write something, test it, see what fails, fix it, test again.

Runtime requirements (pre-flight)

Run this block before any mode that touches scripts (CREATE, IMPROVE, VALIDATE, OPTIMIZE, PACKAGE):

# 1. uv (used by every script in scripts/ and eval-viewer/)
command -v uv >/dev/null || command -v /home/shima/.local/bin/uv >/dev/null \
  || { echo "FAIL: uv not found. Install via curl -LsSf https://astral.sh/uv/install.sh | sh"; exit 1; }

# 2. SKILL_CONDUCTOR_DIR — absolute path to this skill (scripts use relative imports)
SKILL_CONDUCTOR_DIR="$(cd "$(dirname "$0")" 2>/dev/null && pwd || echo ~/.zeroclaw/workspace/skills/skill-conductor)"

# 3. UV_BIN — full path so subprocess and shell tool both work
UV_BIN="$(command -v uv || echo /home/shima/.local/bin/uv)"

# 4. Optional: claude CLI (only Mode 5 OPTIMIZE)
command -v claude >/dev/null || echo "WARN: claude CLI absent — Mode 5 OPTIMIZE will fail"

If uv is absent, stop and tell the user. Don't try to fall back to python3 directly — scripts have inline dependencies (# /// script blocks) that require uv run.

For Mode 1 Step 6 (eval loop) and Mode 2 (improve), the executor subagent needs LLM access. Three options, in order of preference:

1. claude CLI logged in (claude /login or ANTHROPIC_API_KEY in env). Verify: claude --print --model claude-sonnet-4-5 "say ok" returns ok.
2. Anthropic SDK directly via uv run --with anthropic ... if ANTHROPIC_API_KEY is set.
3. On hosts with neither (e.g. zeroclaw with OAuth-only Claude), evals can only be run by the orchestrating agent itself — there's no separate subagent. Mode 1 Step 6 then degrades to: write evals, the user runs them via their own agent, paste outputs back here.

Check the environment up front and tell the user which path applies. Don't pretend to spawn subagents that can't authenticate.

How to communicate

Read context cues. If the user is a skill author iterating on their own work, be direct and technical. If they're new to skills, explain the why behind each step — not just what to do, but why it matters. Default to conversational, not robotic.

• Explain trade-offs when there's a real choice to make
• Use concrete examples over abstract rules
• When something fails, explain the root cause, not just the fix
• Imperative voice in instructions: "Extract the data", not "You should extract"

Modes

Detect mode from context. If ambiguous, ask.

Mode 1: CREATE

Step 1: Capture Intent

Before writing anything, extract 2–3 concrete scenarios.

Ask:

• "What specific task should this skill handle?"
• "What would a user say to trigger it?"
• "What should NOT trigger it?"

Don't move on until you have a clear picture of what the skill does, for whom, and when. This prevents the most common failure: a skill that does something but triggers for the wrong things.

Step 2: Baseline (TDD RED)

Before writing the skill, verify the agent fails without it:

1. Take one scenario from Step 1
2. Run it in a clean session without the skill
3. Document what went wrong — what the agent guessed, what it missed

If the agent already handles it perfectly, the skill is unnecessary. This sounds obvious, but it's the most skipped step and the most valuable one.

Step 3: Architecture

Choose a primary pattern from references/patterns.md (can combine):

Choose degrees of freedom — this determines how much control vs. flexibility the skill gives the agent:

Если скил процедурный (бизнес-процесс с цепочкой действий и решениями: обработка заявки, создание КП, обработка счёта, онбординг, эскалация инцидента) — прочитай references/sop-practices.md. Там 8 принципов из 80-летней SOP-традиции (армия США → авиация → McDonald's): TWI-структура шагов, inline-чеклисты, 5 Why, выбор формата (simple/hierarchical/flowchart), запреты на модальные слова. Это сильно меняет как пишется SKILL.md для процедурных задач.

Step 4: Scaffold

uv run scripts/init_skill.py <skill-name> --path <output-dir> [--resources scripts,references,assets]

Or create manually:

skill-name/
├── SKILL.md          # required — the brain
├── scripts/          # deterministic operations (executed, not loaded)
├── references/       # detailed docs (loaded on demand)
└── assets/           # templates, images for output (never loaded)

Step 5: Write SKILL.md

Frontmatter

---
name: kebab-case-name
description: >
  [Purpose in one sentence]. Use when [triggers].
  Do NOT use for [negative triggers].
---

The description is the single most important line. It determines whether the skill gets triggered at all. Rules:

• name: lowercase, digits, hyphens only. No consecutive hyphens. Matches folder name. Max 64 chars
• description: max 1024 chars. No angle brackets. No process/workflow steps
• Start with purpose, then "Use when...", then "Do NOT use for..."
• Don't put workflow in the description — tested: when the description lists process steps, the agent follows it and skips the body entirely

# GOOD: purpose + triggers, no process
description: Analyze Figma design files for developer handoff. Use when user uploads .fig files or asks for "design specs". Do NOT use for Sketch or Adobe XD.

# BAD: process in description (agent skips body)
description: Exports Figma assets, generates specs, creates Linear tasks, posts to Slack.

Body structure

# Skill Name

## Overview

What this enables. 1-2 sentences. Core principle.

## [Main sections]

Step-by-step with numbered sequences.
Concrete templates over prose.
Imperative voice throughout.

## Common Mistakes

What goes wrong + how to fix.

## Troubleshooting (if applicable)

Error: [message] → Cause: [why] → Fix: [how]

Writing rules

• One term per concept. Pick "template" and stick with it — not template/boilerplate/scaffold
• Progressive disclosure. SKILL.md = brain (<500 lines). References = details. One level deep
• Token budget. Frequently loaded: <200 words. Standard: <500 lines. Heavy: move to references/
• No junk files. No README, CHANGELOG inside the skill
• Scripts: bundle when same code rewritten repeatedly, or operation is fragile. Must return descriptive stdout/stderr on failure
• Imperative voice. Use "Extract the data", not "you should extract" or capitalized "MUST/NEVER" — explanation > rule (see references/sop-practices.md Principle 2)

Step 6: Test Cases & Eval Loop

This is the critical step — most failures hide here. Treat it as three sub-phases.

6a. Pre-flight (before spawning anything)

• evals/evals.json exists with 3–5 prompts (see references/schemas.md)
• Workspace dir created: <skill-name>-workspace/iteration-1/
• Each eval has a descriptive name (not just eval-0) and eval_metadata.json
• Anthropic key for executor subagents is set
• uv and eval-viewer/generate_review.py are reachable from current working dir

If any item fails — fix before proceeding. A missing workspace dir mid-run loses outputs.

6b. Run loop (do all in one turn)

6c. Post-run checklist

• All timing.json files written (one per run)
• Each run has a grading.json with fields text, passed, evidence (not name/met)
• benchmark.json aggregated: uv run scripts/aggregate_benchmark.py <workspace>/iteration-N --skill-name <name>
• Analyst pass done — see agents/analyzer.md for what to look for (non-discriminating assertions, high-variance evals, time/token tradeoffs)
• Eval viewer launched: uv run eval-viewer/generate_review.py <workspace> --skill-name <name> --benchmark <path>
  • In headless mode: --static <output.html> and send file to user
  • For iteration 2+: add --previous-workspace <previous-iteration-path>
• User saw the viewer before I started editing the skill

The last bullet is the trap. If you skip user review and "improve" based on your own reading of outputs, you optimize against your taste, not the user's.

Step 7: Verify & Refactor

1. Does the skill trigger automatically for the right queries?
2. Does the agent follow body instructions (not just description)?
3. Does the output meet use case requirements?
4. Does it NOT trigger on unrelated queries?

If any fail → iterate. Find how the agent rationalizes around the skill, plug loopholes, re-verify.

Mode 2: IMPROVE

Step 1: Diagnose

Read the existing SKILL.md completely. Identify the problem class:

Improvement mindset

1. Generalize from feedback. You're iterating on a few examples, but the skill will be used on thousands of prompts. Don't overfit — avoid fiddly patches or oppressive MUSTs for one test case. Try different metaphors or patterns instead
2. Keep the prompt lean. Read transcripts, not just outputs. If the skill makes the model waste time on unproductive steps, remove those instructions and see what happens
3. Explain the why. LLMs have good theory of mind. Instead of ALWAYS/NEVER in caps, explain the reasoning — it's more powerful and robust. If you're writing rigid rules, reframe as explanations
4. Look for repeated work. If all test runs independently write the same helper script, bundle it in scripts/. Saves every future invocation from reinventing the wheel
5. For procedural skills - apply SOP practices. If improving a process skill (handling a ticket, generating a quote, escalating an incident), read references/sop-practices.md. The 8 principles from SOP tradition (TWI step structure, inline checklists, 5 Why, removing modal weasel words like "regularly/typically/как правило") map directly to skill failure modes — silent improvisation, missed edge cases, agents skipping checklists at end of doc

Step 2: Eval Iteration Loop

The improvement cycle mirrors CREATE Step 6, but focused on the broken behavior:

1. Run the failing case with current skill → document failure
2. Apply fix using writing rules from CREATE Step 5
3. Run eval again → grade with agents/grader.md
4. Launch viewer: uv run eval-viewer/generate_review.py <workspace>
   • Headless/Cowork: use --static <output.html> instead of live server
5. Review, provide feedback, iterate

Step 3: Blind Comparison (optional, for major changes)

When you have two meaningfully different versions:

1. Run both versions on the same evals
2. Spawn agents/comparator.md — receives outputs A and B without knowing which skill produced which
3. Comparator scores on rubric (content + structure, 1–5 each) and picks a winner
4. Spawn agents/analyzer.md — unblinds results, analyzes WHY the winner won
5. Apply insights to improve the losing version

This prevents bias. The comparator judges output quality, not skill design.

Mode 3: VALIDATE

Three stages, run in order.

Stage 1: Structural Validation

uv run scripts/eval_skill.py <skill-folder>

Checks: frontmatter, naming, description quality, process leak detection, body size, structure, scripts. Target: 10/10, no warnings.

Stage 2: Discovery (trigger testing)

Generate 6 test prompts:

• 3 that SHOULD trigger the skill
• 3 that should NOT (similar-sounding but wrong domain)

Run each in clean session. Target: 6/6 correct.

For automated trigger testing at scale, use:

uv run scripts/run_eval.py --eval-set <path> --skill-path <path> --runs-per-query 3

Stage 3: 5-Axis Scoring

Rate on 5 axes (1–10 each):

Interpretation: 45–50 production ready · 35–44 solid · 25–34 needs work · <25 rewrite

Mode 4: REVIEW

Quick quality gate for third-party skills.

Checklist (pass/fail)

[ ] SKILL.md exists, exact case
[ ] Valid YAML frontmatter (name + description)
[ ] name: kebab-case, matches folder, ≤64 chars
[ ] description: ≤1024 chars, no angle brackets
[ ] description has triggers ("Use when...")
[ ] description has NO workflow/process steps
[ ] No README.md inside skill folder
[ ] SKILL.md < 500 lines
[ ] References max 1 level deep
[ ] Scripts tested and executable
[ ] No hardcoded paths/tokens/secrets

Then run VALIDATE Stage 2 (discovery) on the description. Report score + checklist.

The checklist exists because these are the failure modes that actually happen in practice — especially process-in-description, which causes the agent to skip the body entirely.

Mode 5: OPTIMIZE

Automated description optimization. The description competes with other skills for Claude's attention — optimization finds the wording that triggers most accurately.

How it works

1. Create an eval set: 20 queries (10 should-trigger, 10 should-not)

Writing good eval queries

Queries must be realistic — concrete, detailed, with file paths, context, abbreviations, typos. Not "Format this data" but "my boss sent Q4 sales final FINAL v2.xlsx, add profit margin % column, revenue is col C costs col D".

Should-trigger (10): Different phrasings of the same intent — formal, casual, implicit. Include cases where user doesn't name the skill but clearly needs it. Add competing-skill edge cases.

Should-NOT-trigger (10): Near-misses that share keywords but need something different. Adjacent domains, ambiguous phrasing. "Write fibonacci" as negative for PDF skill = useless — too easy. Make negatives genuinely tricky.

Triggering mechanics: Claude only consults skills for tasks it can't handle directly. Simple queries ("read this PDF") won't trigger skills regardless of description — Claude handles them with basic tools. Eval queries must be substantive enough that consulting a skill would help.

2. Review queries in the browser: assets/eval_review.html
3. Run the optimization loop:

uv run scripts/run_loop.py \
  --eval-set evals/eval_set.json \
  --skill-path <skill-dir> \
  --model claude-sonnet-4-20250514 \
  --max-iterations 5 \
  --holdout 0.4 \
  --verbose

The loop:

• Splits queries into train (60%) and test (40%) to prevent overfitting
• Each iteration: evaluates current description → Claude proposes improvement → re-evaluates
• Improvement model sees only train results (blinded to test)
• Selects the best description by test score
• Opens live HTML report automatically

Supporting scripts

Mode 6: PACKAGE

1. Run REVIEW checklist (Mode 4)
2. Validate:

uv run scripts/quick_validate.py <skill-folder>

3. Package:

uv run scripts/package_skill.py <skill-folder> [output-dir]

Creates skill-name.skill (zip with .skill extension). Verify: unzip in temp dir, check structure intact.

Quick Reference

Skill categories

1. Document/Asset Creation — consistent output (docs, designs, code)
2. Workflow Automation — multi-step processes with methodology
3. MCP Enhancement — workflow guidance on top of tool access
4. Procedural / Process — business procedures with decision points and exceptions (handling a request, generating a quote, processing an invoice, onboarding, escalation). For these → read references/sop-practices.md

File purposes

Progressive disclosure budget

Description formula

[What it does] + Use when [triggers, file types, symptoms]. + Do NOT use for [negatives].

Reference Files