GEO Prompt Architecture

v0.1.3

Use when the user wants to generate, structure, score, or audit GEO monitoring prompts for a client. Trigger when building topic-first prompt sets from a web...

0· 117·0 current·0 all-time
byTim@geo-seo
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 (GEO Prompt Architecture) match the contents: prompt-generation, topic maps, scoring and audit guidance. The repo is instruction-only and includes schemas, examples and references that are relevant to the stated purpose. Nothing in the manifest requests unrelated credentials, binaries, or platform access.
Instruction Scope
SKILL.md instructs the agent to reconstruct client models, build topic maps, and generate structured prompts. The instructions operate on pasted briefs, URLs, screenshots, and provided inputs; they do not direct reading of arbitrary system files, environment variables, or other skills' secrets. The guidance is prescriptive (schemas, required fields) rather than open-ended data collection.
Install Mechanism
No install spec and no code files that would be written to disk at install time. Instruction-only skills have lower risk because they do not pull external binaries or archives.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. SKILL.md mentions that live browsing is helpful but does not require credentials — this is a reasonable optional capability for validating public websites; it does not demand unrelated secrets.
Persistence & Privilege
always is false and the skill does not request persistent system-level privileges or to modify other skills. Autonomous invocation is allowed (platform default) but there is no evidence of privilege escalation or long-term persistence behavior in the repo.
Assessment
This skill is instruction-only and appears coherent for building GEO/topic-first prompt packs. Before using it: (1) avoid pasting private analytics, Search Console exports, CRM dumps, or any credentials unless you trust the agent and environment; the skill can work from public website URLs or pasted briefs. (2) If you enable any browsing/file-access tools for the agent, be aware the agent may fetch public website content to infer topics — grant those tools only if acceptable. (3) Review generated prompt sets for brand or privacy-sensitive content before sharing externally. The repo contains helpful schemas and examples; there are no hidden endpoints or env-var requests in the provided files.

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

latestvk975gvfk909h92rg9qkkajq2ps832xks

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Comments