paseo-advisor
Spin up a single agent as an advisor — second opinion on the current task. Use when the user says "advisor", "second opinion", "what does X think", or wants an outside take without delegating the work itself.
Security Assessment
About paseo-advisor
Paseo Advisor is a lightweight orchestration skill for the Paseo agent platform that spins up a single "advisor" agent to give a second opinion on the task at hand. Rather than delegating the work itself, it produces an outside judgment: the advisor reads the situation, offers a recommendation with reasoning, and hands the decision back to you. It solves the common problem of wanting a fresh perspective — a design critique, an audit, or root-cause analysis — without handing off ownership of the work.
The workflow depends on the companion Paseo skill. It calls `list_profiles`, then selects an advisor either from a user-named profile (via `--profile`) or by matching the question to a profile's notes, deliberately preferring a different provider family for genuine contrast. It assembles a self-contained briefing (the sharp question, what was considered and ruled out, relevant file paths, and an explicit ask for a recommendation) and always appends an analysis-only suffix instructing the advisor not to edit, create, or delete files. It can also forward another skill (for example `/unslop` or `/diagnose`) so the advisor runs that skill against the current task, and it supports persistent advisors for ongoing input across several decisions.
This skill is aimed at developers and agent users working inside the Paseo ecosystem who want a structured, read-only second opinion during a task. Typical use cases include design reviews, build-failure diagnosis, code-diff critiques, and any decision point where a contrasting viewpoint helps before committing to an approach.
FAQ
What does the Paseo Advisor skill actually do?
It creates a single advisor agent that reads your current situation and returns a judgment or recommendation with reasoning. You stay the orchestrator and decide what to do with the advice — the advisor never drives or performs the work.
What do I need to use it?
You must first read the companion paseo skill, since the advisor is materialized through Paseo's create_agent and list_profiles. Advisors are chosen from configured profiles and their notes.
Can the advisor change my files or write code?
No. Every briefing ends with an explicit analysis-only instruction telling the advisor not to edit, create, or delete any files and not to write code. It is designed as a read-only second opinion.
How do I pick which advisor responds?
Name a profile with --profile, or let the skill match the profile whose notes best fit the question. When several fit, it prefers a different provider family from your own so the second opinion is genuinely fresh.
Can I keep the same advisor for multiple decisions?
Yes. If you want ongoing input, the advisor is not archived after the first reply; you send follow-ups as needed and archive when you are done or when the topic shifts enough to warrant fresh context.
Install paseo-advisor
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
getpaseo/paseo