ClawHub

Arquitecto Categórico

v1.0.1

Disena y audita arquitecturas de datos y APIs con teoria de categorias. Usar cuando haga falta formalizar un dominio en PostgreSQL DDL, JSON Schema, OpenAPI,...

⭐ 0· 124·0 current·0 all-time
by@felix-antonio-sl
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description ask for category-theory modelling and audits; the bundle contains many domain-specific playbooks and KB files that the SKILL.md references. No unrelated env vars, binaries, or install steps are requested — the resources are proportional to the stated purpose.
✓
Instruction Scope
Runtime instructions direct the agent to classify the request, load specific local playbooks under {baseDir}/references (engine-map.md, kb-map.md, etc.), ask clarifying questions when a modelling tension exists, and emit declarative artefacts (DDL/JSON Schema/OpenAPI/diagrams). The instructions do not ask to read arbitrary system files, environment secrets, or send data to external endpoints; they explicitly treat the corpus as read-only and limit loaded files to 1–3 per turn.
✓
Install Mechanism
There is no install spec and no code files to execute; this is an instruction-only skill (lowest install risk). All supporting files are local documentation/playbooks included in the bundle.
✓
Credentials
The skill declares no required environment variables, credentials, or config paths. The instructions reference only files packaged in the skill under {baseDir}/references, which is consistent with its purpose.
✓
Persistence & Privilege
Flags: always=false and normal autonomous invocation allowed. The skill does not request long-term presence, does not modify other skills or global agent configuration, and does not attempt to store or escalate privileges.
Assessment
This skill appears coherent and self-contained: it reads the included reference playbooks and emits declarative artefacts (DDL, JSON Schema, OpenAPI, diagrams) using category-theoretic guidance. Before installing: (1) confirm you trust the skill bundle files (they will be read by the agent), (2) avoid feeding secrets or private credentials to the skill (it does not need them), and (3) review any generated DDL or migration recommendations before applying them to production systems. If you want stricter control, limit autonomous invocation or review outputs interactively rather than allowing fully autonomous changes.

Like a lobster shell, security has layers — review code before you run it.

latestvk977aaxamm8vw0vts85z2p9ec183yqws

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments