gemini-task-contextualizer

Generates structured briefing prompts for Gemini explorer and worker subagents.

2|1|Updated Jun 25, 2026
One-click install
npx skills add https://github.com/Consiliency/agent-harness --skill gemini-task-contextualizer-consiliency
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: gemini-task-contextualizer
Source: https://github.com/Consiliency/agent-harness/tree/main/skills-src/gemini/gemini-task-contextualizer
Command: npx skills add https://github.com/Consiliency/agent-harness --skill gemini-task-contextualizer-consiliency

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Delegating work to Gemini subagents with vague prompts leads to wasted effort, conflicting edits, and missing context. This Skill standardizes subagent briefs so every delegation includes goals, file paths, scope boundaries, and ownership rules. ## Core Features & Use Cases - Mandatory Brief Structure: Enforces a checklist covering goal, starting files, architecture context, scope boundary, ownership, related files, expected output, and a non-reversion coordination rule. - Explorer Briefs: Produces read-only reconnaissance prompts that demand file:line evidence for research questions. - Worker Briefs: Produces implementation prompts with disjoint write ownership, task lists, and changed-file reporting requirements. - Use Case: Before spawning three parallel Gemini agents to investigate a codebase and implement disjoint features, generate a contextualized brief for each so they avoid overwriting each other's edits. ## Quick Start Ask the assistant to prepare a Gemini subagent brief for delegating a bounded implementation task with explicit owned files and scope boundaries.

Frequently Asked Questions about gemini-task-contextualizer

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

FAQPage Schema
How do I write a good prompt for a Gemini coding subagent?▼

Include a one-sentence goal, exact starting file paths, brief architecture context, scope boundaries, file ownership rules, and an expected output format. Add a coordination rule telling the agent not to revert edits made by others.

How to run parallel AI agents without conflicting edits?▼

Assign each worker agent disjoint write ownership over specific files or globs, and mark shared files as read-only. Include a rule that agents must adjust to peer changes instead of overwriting them.

When should I not delegate a task to a subagent?▼

Do not spawn a subagent just because a task is large. If the next critical-path step depends on the answer, do the work locally instead of adding delegation overhead.

What is the difference between an explorer and a worker subagent brief?▼

Explorer briefs are read-only reconnaissance prompts that require file:line evidence for findings. Worker briefs grant write ownership of specific files and require a report of changed paths, tests run, and blockers.

Why do AI subagents revert each other's changes?▼

Subagents lack awareness of concurrent edits unless told otherwise. Adding an explicit coordination rule stating the agent is not alone in the codebase prevents it from overwriting peer work.