🎯

no-task-output

🎯Skill

from parcadei/continuous-claude-v3

What it does

Prevents context flooding by avoiding TaskOutput and using synchronous Task mode instead.

no-task-output

Installation

Install skill:

npx skills add https://github.com/parcadei/continuous-claude-v3 --skill no-task-output

Last UpdatedJan 26, 2026

View on GitHub Back to Skills

Skill Details

SKILL.md

Never Use TaskOutput

Overview

# Never Use TaskOutput

TaskOutput floods the main context window with agent transcripts (70k+ tokens).

Rule

NEVER use TaskOutput tool. Use Task tool with synchronous mode instead.

Why

TaskOutput reads full agent transcript into context
This causes mid-conversation compaction
Defeats the purpose of agent context isolation

Pattern

```

# WRONG - floods context

Task(run_in_background=true)

TaskOutput(task_id="...") // 70k tokens dumped

# RIGHT - isolated context, returns summary

Task(run_in_background=false) // Agent runs, returns summary

```

Source

Session where TaskOutput caused context overflow

More from this repository10

🎯

agentica-claude-proxy🎯Skill

Enables seamless integration between Agentica agents and Claude Code CLI by managing proxy configurations, tool permissions, and response formatting.

🎯

git-commits🎯Skill

Manages git commits by removing Claude attribution, generating reasoning documentation, and ensuring clean commit workflows.

🎯

debug-hooks🎯Skill

Systematically diagnose and resolve hook registration, execution, and output issues in Claude Code projects by checking cache, settings, files, and manual testing.

🎯

migrate🎯Skill

Systematically researches, analyzes, plans, implements, and reviews migrations across frameworks, languages, and infrastructure with minimal risk.

🎯

background-agent-pings🎯Skill

Enables background agent execution with system-triggered progress notifications, avoiding manual polling and context flooding.

🎯

agentica-infrastructure🎯Skill

Provides comprehensive reference and infrastructure for building sophisticated multi-agent coordination patterns and workflows with precise API specifications and tracking mechanisms.

🎯

system-overview🎯Skill

Generates a comprehensive summary of the current system's configuration, components, and key metrics across skills, agents, hooks, and other core systems.

🎯

cli-reference🎯Skill

Provides comprehensive CLI commands and flags for interacting with Claude Code, enabling headless mode, automation, and session management.

🎯

braintrust-tracing🎯Skill

Traces and correlates Claude Code session events across parent and sub-agent interactions using comprehensive Braintrust instrumentation.

🎯

morph-apply🎯Skill

Rapidly edits files using AI-powered Morph Apply API with high accuracy and speed, without requiring full file context.