role

Activates, rotates, and deactivates session roles that apply role-specific lenses to workflow phases.

Updated May 27, 2026
One-click install
npx skills add https://github.com/jokerman89/lintel --skill role-jokerman89
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: role
Source: https://github.com/jokerman89/lintel/tree/main/skills/role
Command: npx skills add https://github.com/jokerman89/lintel --skill role-jokerman89

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams working with AI agents often need the agent to adopt a specific professional perspective (such as a product manager or security reviewer) during a session, but switching context manually is inconsistent and loses continuity. This Skill manages the full lifecycle of session roles so the agent applies the right perspective at the right time. ## Core Features & Use Cases - Role lifecycle management: Activate a role with a lightweight ~500-token load, deactivate it, or atomically rotate to a different role mid-session while preserving session memory. - Artifact framing: Apply the active role's decision criteria, cold knowledge, and voice to review an artifact such as a design doc, proposal, or PR description, with prioritized P1-P3 edit suggestions. - On-demand deep dives: Load the complete role file (~2-3k tokens) only when full expertise is needed, with token cost reporting and private-role sensitivity gates. - Use Case: During a planning session, activate a product-manager role to frame a specification against customer outcomes, then rotate to an engineering role before implementation begins. ## Quick Start Ask the agent to activate the product-manager role for this session and apply its lens to the current specification document.

Frequently Asked Questions about role

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

FAQPage Schema
How do I activate a role for an AI agent session?▼

Run /li:role <role-id> to activate a role with a lightweight ~500-token load of its identity, voice, and outcome lens. The role is recorded in the profile so subsequent workflow phases apply its overlay automatically.

How do I switch roles mid-session without losing context?▼

Use /li:role --rotate <role-id> to atomically swap roles. The new role file is verified before the current one is deactivated, and session memory such as lessons and artifacts is preserved.

Can I apply a role's perspective to an existing document?▼

Yes, /li:role --frame <artifact> applies the active role's decision criteria, cold knowledge, and voice to an artifact like a design doc or PR description. It surfaces strengths, gaps, and prioritized P1-P3 edit suggestions for operator approval.

What is the difference between role activation and deep-dive?▼

Activation loads a ~500-token summary suitable for session start, while --deep-dive loads the full 2-3k token role file including cold knowledge and decision criteria. Deep-dive reports token cost and is intended for high-stakes work, not session start.

How are private roles handled differently from public ones?▼

Private roles require confirmation before activation or deep-dive, and their framing findings are written to gitignored role-lens-notes files instead of inline edits. Sensitive context is never propagated into public artifacts without explicit per-edit approval.