Menü
Use when explaining what each file, script, reference, asset, or template does inside a skill package.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
Demonstrates the Agent Skill Anatomy harness with a tiny local skill.
Use when selecting and composing skill anchors into a chosen reuse form without assuming a full workflow is required.
Use when auditing whether a skill decomposition is evidence-grounded, over-inference-safe, and publishable.
Use when comparing multiple model outputs for the same skill decomposition task across quality, omissions, and fit.
Use when turning structured skill decomposition artifacts into beginner, expert, report, or Obsidian reading layers.
Use when extracting reusable patterns, templates, checklists, anti-patterns, and transfer rules from analyzed skills.
| name | asa-resource-role-analyzer |
| description | Use when explaining what each file, script, reference, asset, or template does inside a skill package. |
| internal_meta_skill | true |
| asa_role | structure_analyst |
| output_contract | structure_analysis.resource_roles |
Use this internal meta-skill when a model must explain what each important file or resource does in a skill package.
This skill fixes the failure mode where analysis lists files but does not explain whether they are instructions, templates, executable tools, references, examples, assets, tests, configs, schemas, or compliance material.
SKILL.md content and package layout.scripts, references, assets, examples, templates, tests, and root files.Return resource role content suitable for structure_analysis.json:
resource_roles:
- path:
role:
stage:
read_policy: must_read | on_demand | optional | ignore
reuse_value: high | medium | low | unknown
risk:
evidence: []
In addition to resource role content, emit minimal anchor cards for resources that may affect reuse or composition. Keep this lightweight; do not copy files or decide whether to apply them to a user project.
Recommended anchor types:
resource_anchor: an important file, directory, reference, asset, or config.template_anchor: a reusable scaffold, prompt template, UI template, document template, or schema.tool_anchor: a script, command, API helper, or executable utility.dependency_anchor: package, runtime, environment variable, service, or external requirement.portable_resource_anchor: a resource that can be copied or adapted safely.non_portable_resource_anchor: a resource that should stay reference-only or requires permission/adaptation.Minimal anchor card shape:
anchors:
- id: anchor.<skill_id>.resource.<slug>
source_skill_id:
anchor_type: resource_anchor
name:
summary:
evidence:
path:
quote:
confidence: low | medium | high
reuse_modes:
- reference_only
- template_pack
- embedded_in_existing_workflow
risks:
- missing dependency, license risk, context size, or runtime assumption
Anchor rules:
template_anchor when they shape downstream outputs.tool_anchor only when execution is supported by source evidence or dependency analysis.dependency_anchor.portable_resource_anchor only when copying/adapting appears safe and useful.non_portable_resource_anchor when a resource is license-bound, context-bound, environment-bound, secret-bearing, or too coupled to the original skill.asa-anchor-composition-planner.structural evidence for file path and layout claims.explicit evidence when SKILL.md says to read or use a file.inferred evidence when assigning likely stage from filename or package convention.A strong resource analysis lets a reader answer: