Skip to content

Auto-dispatch skill: questio_gap → pipeio_run

Problem

questio_gap identifies the highest-impact unblocked milestone and recommends which pipeline to run. But the agent still has to manually resolve the recommendation into a pipeio_run call. This is a composable skill, not new infrastructure.

Proposed skill: questio-dispatch

Workflow: 1. Call questio_gap(question_id) → get actionable milestones + pipeline names 2. For each actionable milestone, resolve pipeline → flow name 3. Call pipeio_flow_status(flow) to check if flow is configured and ready 4. Call pipeio_target_paths(flow) to get concrete targets 5. Call pipeio_run(targets) to execute 6. After run completes, prompt the agent to inspect output and create a result note via questio-record

Example session

Agent: questio_gap("H1")
  → "Root blocker: ttl-removal-validated. Pipeline: preprocess_ieeg. Actionable."

Agent: questio-dispatch("H1")
  → Resolves: preprocess_ieeg flow, ttl mod
  → pipeio_flow_status("preprocess_ieeg") → configured, dry-run passes
  → pipeio_run(targets=[...]) → executes
  → "Pipeline complete. Inspect output and record results? [Y/n]"

Scope

This is a prompt-based skill composing existing MCP tools. No new tools needed. Difficulty: low. The skill template already exists in questio-session — extract the dispatch loop into a standalone skill.

Boundary with questio-session

questio-session is the full research session (orient → plan → ground → execute → record → close). questio-dispatch is just the execute step — called within a session or standalone when the agent already knows what to run.

Source context: pixecog

PixEcog (pixecog): Neuropixels and ECoG dataset and analysis

Recent commits: 

36f9326 Add result note directory and sample note
62841d9 Add questio YAML data model (questions.yml + milestones.yml)
9b2f6fa Scaffold ecephys TTL removal mod, flow overview + mod docs, demo notebook

README:

type: readme

Quick Start for Collaborators

Follow this checklist to get started with Pixecog documentation and workflows.

🐀 Pixecog Project — Compact Overview

Core principles

  • One immutable BIDS raw dataset (raw/) as the canonical baseline
  • Each analysis pipeline ha