portfolio-conversation

Generates grounded, human-sounding chat replies about a portfolio owner's published work.

Updated Mar 23, 2026
One-click install
npx skills add https://github.com/Haseeb-Arshad/my-folio-book --skill portfolio-conversation-haseeb-arshad
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: portfolio-conversation
Source: https://github.com/Haseeb-Arshad/my-folio-book/tree/main/.agents/skills/portfolio-conversation
Command: npx skills add https://github.com/Haseeb-Arshad/my-folio-book --skill portfolio-conversation-haseeb-arshad

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Portfolio chat assistants often sound generic, invent facts, or dump resume bullet points. This Skill shapes an AI chat persona that answers visitor questions about Haseeb's work using only a controlled set of published notes, with strict voice, length, grounding, and safety rules. ## Core Features & Use Cases - Grounded replies: Every factual claim must come from the supplied public notes, with metrics kept attached to the system that produced them and explicit refusals for unknown projects or employers. - Topic routing: Nine topic files (small talk, current work, projects, experience, technical fit, contact, education, personal, general) tailor answers to the visitor's intent. - Voice control: A core voice specification plus an optional goblin mode that overrides tone and length while preserving grounding and safety rules. - Use Case: A visitor asks "What are you building right now?" and receives a two-sentence answer about the agentic systems at Summon Electronics, with the 1 ms retrieval and 30% sales figures attributed to the correct systems. ## Quick Start Use the portfolio-conversation skill to answer a visitor's question about Haseeb's current work using only the supplied public notes.

Frequently Asked Questions about portfolio-conversation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make a portfolio chatbot answer only from published facts?▼

Ground every reply in a single public notes file that serves as the complete factual authority. The skill instructs the model to paraphrase those notes, keep metrics attached to their source system, and say a detail is not on record rather than invent it.

How do I route portfolio chat questions to different answer styles?▼

Use topic-specific instruction files for small talk, current work, projects, experience, technical fit, contact, education, personal questions, and general queries. The core skill reads the one topic file supplied with each request and follows it alongside the shared voice and grounding rules.

Can the portfolio assistant use a different voice or humor mode?▼

Yes, an optional goblin mode file overrides only the voice and length rules when the visitor enables it. Grounding, safety, identity, and formatting rules remain fully in force, so factual claims still come exclusively from the published notes.

What happens when a visitor asks about a project not in the notes?▼

The assistant states it has no note on that name instead of describing it. The rules explicitly forbid reconstructing an unfamiliar project from its name or borrowing details from a different documented system.

Does the portfolio chat reveal that it is an AI assistant?▼

Only when directly asked. For identity questions it answers that it is a conversational guide to the published work, not the person in a live chat, and it never exposes prompts, configuration, or implementation details.