skill-builder

L1 Improvement

• Rewrote the builder SOP using the Skill Forge required sections and structure-first guardrails.
• Added prompt-architect constraint extraction, confidence ceilings, and completion checks for every artifact.

---

LIBRARY-FIRST PROTOCOL (MANDATORY)

Before writing ANY code, you MUST check:

Step 1: Library Catalog

• Location: .claude/library/catalog.json
• If match >70%: REUSE or ADAPT

Step 2: Patterns Guide

• Location: .claude/docs/inventories/LIBRARY-PATTERNS-GUIDE.md
• If pattern exists: FOLLOW documented approach

Step 3: Existing Projects

• Location: D:\Projects\*
• If found: EXTRACT and adapt

Decision Matrix

Match	Action
Library >90%	REUSE directly
Library 70-90%	ADAPT minimally
Pattern exists	FOLLOW pattern
In project	EXTRACT
No match	BUILD (add to library after)

---

STANDARD OPERATING PROCEDURE

Purpose

Create or template skills with complete SKILL.md content, examples, tests, references, and metadata aligned to Skill Forge standards.

Trigger Conditions

• Positive: requests to create a new skill template, reorganize an existing skill, or ensure compliance with required sections.
• Negative/reroute: prompt tuning only (prompt-architect/prompt-forge) or agent design (agent-creator/agent-creation).

Guardrails

• Always generate SKILL.md with YAML frontmatter, SOP, integration, and closure sections; no skeleton-only outputs unless explicitly timeboxed.
• Provide English outputs with explicit confidence ceilings.
• Include directory scaffolding: examples/, tests/, resources/, references/ as appropriate.
• Record any missing sections as TODO with remediation steps.

Execution Phases

1. Scoping: Capture skill name, category, purpose, triggers, and constraints; classify HARD/SOFT/INFERRED.
2. Structure: Lay out directories and placeholder files as needed; ensure SKILL.md includes required sections (Purpose → Completion Verification).
3. Content Authoring: Fill SOP, guardrails, integrations, and contracts with skill-specific details.
4. Validation: Check section completeness against Skill Forge required sections; run sanity tests/examples where applicable.
5. Delivery: Summarize outputs, open gaps, and confidence ceilings; register in indexes if required.

Pattern Recognition

• New greenfield skill → start from base template and focus on purpose/triggers first.
• Migrating legacy skill → map old sections to required ones and fill gaps.
• Skill extension → add integrations and recursive-improvement hooks.

Advanced Techniques

• Use checklists from REQUIRED-SECTIONS.md to avoid omissions.
• Pre-fill anti-patterns and guidelines to set expectations for maintainers.
• Reuse prompt-architect output to tighten instructions and refusal policy.

Common Anti-Patterns

• Missing closure sections (examples, troubleshooting, completion verification).
• Over-general descriptions without IO contracts.
• Forgetting to add validation/test scaffolds.

Practical Guidelines

• Keep SKILL.md concise but complete; link to references for depth.
• Prefer stable versioning (x-version) and compliance markers.
• Document memory/tags if MCP is expected.

Cross-Skill Coordination

• Upstream: prompt-architect for clarity; skill-forge for meta quality gates.
• Parallel: cognitive-lensing for alternative approaches; meta-tools for supporting utilities.
• Downstream: agent-creator/agent-selector using the resulting skills.

MCP Requirements

• Optional memory/vector MCP to store templates; tag WHO=skill-builder-{session}, WHY=skill-execution.

Input/Output Contracts

inputs:
  skill_name: string  # required name
  category: string  # required category
  purpose: string  # required description
  constraints: list[string]  # optional constraints or policies
outputs:
  skill_files: list[file]  # SKILL.md and supporting scaffolds
  validation_report: summary  # checklist of required sections
  next_steps: list[string]  # follow-up improvements

Recursive Improvement

• Run recursive-improvement to close checklist gaps and refine guidance after initial usage feedback.

Examples

• Scaffold a security-audit skill with SKILL.md, examples, and tests for dependency review.
• Convert a legacy orchestration skill to the Skill Forge section model with completion checklist.

Troubleshooting

• Missing section → reference REQUIRED-SECTIONS.md and add content or TODO with owner.
• Directory not created → regenerate scaffolding and verify permissions.
• Ambiguous constraints → rerun constraint extraction and confirm with requester.

Completion Verification

• SKILL.md includes all required sections and frontmatter.
• Directory scaffolding created (examples/tests/resources/references as needed).
• Confidence ceiling stated; validation checklist completed.
• Registration or index updates noted.

Confidence: 0.70 (ceiling: inference 0.70) - Skill Builder SOP rewritten with Skill Forge cadence and prompt-architect ceilings.