Ai Productivity

High-speed intake for shaping vague prompts, triaging complex tasks, and compressing context for efficient execution.

Install
cmdop skills install agensi-ai-productivity

AI Productivity is a skill published by Agensi that acts as an intake layer for agentic workflows. It addresses three recurring problems: vague or underspecified prompts, context bloat from long session logs, and high-risk requests that lack clear success criteria.

The skill supports four defined workflows. Request Triage scopes broad or risky tasks before any execution begins, catching ambiguity that would otherwise cause an agent to drift. Context Compression distills long session histories into a structured summary of facts, decisions, and next steps, reducing the token load passed to downstream agents or tools. Lightweight Rewriting takes loosely stated ideas and converts them into structured prompts that include constraints and formatting rules. Agent Handoffs produce standardized briefs formatted for consumption by multi-agent systems using a multi-agent-coordinator.

Output format depends on the input: a direct answer, a structured internal brief with Summary, Decisions, and Open Questions sections, or a refined prompt ready for immediate execution. No environment variables are required and no external services are listed in the facts.

This skill is appropriate when an orchestration layer needs consistent, logic-driven request shaping before delegating work. It is not a data integration tool, does not connect to databases or APIs, and provides no tools for querying external systems.

Use cases

  • Triage a broad or ambiguous agent task before execution to prevent drift
  • Compress a long session log into a concise brief of facts, decisions, and open questions
  • Rewrite a vague idea into a structured prompt with explicit success criteria
  • Generate a standardized handoff brief for a multi-agent coordinator
  • Strip irrelevant context from a prompt to reduce token consumption
  • Enforce consistent request shaping logic across multiple agentic workflows

When to use it

  • When agent tasks are frequently underspecified and cause execution errors
  • When session context is growing too large to pass efficiently to downstream agents
  • When handing off work between specialized agents that need a clean, structured brief
  • When consistent prompt-shaping logic needs to be applied across many requests

When not to use it

  • When the task requires querying or writing to a database such as Postgres
  • When integration with external APIs or data sources is needed
  • When the workflow requires tool invocations — this skill exposes no tools
  • When a transport protocol or environment-variable-based configuration is required