asa-reuse-pattern-miner

name	asa-reuse-pattern-miner
description	Use when extracting reusable patterns, templates, checklists, anti-patterns, and transfer rules from analyzed skills.
internal_meta_skill	true
asa_role	pattern_miner
output_contract	patterns.reuse_assets + letuen.reuse_anchors

ASA Reuse Pattern Miner

Use this internal meta-skill when a model must convert one or more analyzed skills into reusable design assets.

This skill fixes the failure mode where pattern mining becomes a loose summary instead of reusable anchors, composition candidates, anti-patterns, and optional solidification templates that another author can apply.

Inputs

• Skill summaries and source paths.
• Structure analysis, workflow analysis, and reviewer outputs.
• Resource roles, workflow steps, evidence strength, and reusable artifacts.
• Optional model comparison notes when available.

Process

1. Identify mechanisms that can transfer beyond one repository.
2. Separate concrete implementation details from reusable design principles.
3. For each candidate pattern, write problem, solution, when-to-use, counterexample, and evidence.
4. Mark pattern maturity as candidate, emerging, or established based on example count and mechanism similarity.
5. Extract checklists only when steps are actionable.
6. Extract templates only when placeholders and usage boundaries are clear.
7. Record anti-patterns when the source exposes a likely failure mode.
8. Convert transferable mechanisms into reuse, composition-candidate, anti-pattern, and solidification-template anchors.
9. Avoid copying source text; abstract the method and point back to evidence.

Output Contract

Return reusable assets suitable for patterns.json and Obsidian notes:

reuse_assets:
  patterns: []
  templates: []
  checklists: []
  anti_patterns: []
  extension_ideas: []
  anchors:
    reuse_anchor:
      - id:
        name:
        problem:
        reusable_mechanism:
        when_to_use: []
        when_not_to_use: []
        source_patterns: []
        reuse_modes:
          - reference_only
          - temporary_composition
          - prompt_patch
          - workflow_step
          - learning_note
        portability: high | medium | low
        confidence: high | medium | low | unknown
        evidence: []
    composition_candidate_anchor:
      - id:
        name:
        selected_anchors: []
        composition_form: reference_only | temporary_composition | prompt_patch | agent_rule_patch | workflow_insert | tool_adapter | template_pack | quality_gate | learning_note | full_workflow_blueprint | new_skill_spec
        user_goal:
        expected_output:
        required_adapters: []
        conflict_risks: []
        preserve_existing_skills: true
        solidification_required: true | false
        confidence: high | medium | low | unknown
        evidence: []
    anti_pattern_anchor:
      - id:
        name:
        failure_mode:
        appears_when: []
        avoid_by: []
        affected_anchor_types: []
        severity: high | medium | low
        evidence: []
    solidification_template_anchor:
      - id:
        name:
        target_form: full_workflow_blueprint | new_skill_spec | template_pack | quality_gate
        use_only_when: []
        required_sections: []
        required_anchors: []
        optional_anchors: []
        non_destructive_output:
        rollback_note:
        evidence: []

Anchor Output Rules

Emit anchors that let users split, borrow, combine, or optionally solidify useful parts of a skill without forcing a full workflow.

• reuse_anchor: a transferable mechanism that can stand alone as a note, prompt patch, checklist, or workflow step.
• composition_candidate_anchor: a suggested combination of anchors for a concrete user goal, with composition form and conflict risks.
• anti_pattern_anchor: a reusable warning that prevents bad recombination, overgeneralization, or unsafe copying.
• solidification_template_anchor: an optional template for turning selected anchors into a larger workflow, quality gate, template pack, or new skill.

A composition candidate must state whether solidification is required. Default to solidification_required: false unless the selected goal cannot work as a temporary composition. Preserve existing user skills by default and represent conflicts as risks or adapters, not overwrites.

Evidence Rules

• Every pattern example should include skill_id, source_path, and evidence when available.
• A one-skill pattern is only a candidate unless there is strong external evidence.
• Similar names are not enough; compare mechanism, workflow role, and resource usage.
• If reviewer confidence is low, downgrade pattern confidence.

Failure Modes

• Do not overgeneralize from one skill.
• Do not create generic advice that cannot be applied.
• Do not call a pattern established without multiple matching examples.
• Do not copy examples as templates without explaining how to adapt them.

Quality Rubric

A strong reuse pattern asset answers:

• What problem does this pattern solve?
• What mechanism makes it reusable?
• When should it not be used?
• What evidence shows the mechanism works in the analyzed skill?