skills.rest
bthillerup
bthillerupCOMMUNITY
Explore all skills by this author →
GitHub

filesystem-context

Offload large tool outputs to files for dynamic context discovery.

1|Updated Feb 4, 2026
Search Tags:#context management#filesystem#agent memory#context window#dynamic loading#tool output
One-click install
npx skills add https://github.com/bthillerup/bens-garage-session-2 --skill filesystem-context-bthillerup
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: filesystem-context
Source: https://github.com/bthillerup/bens-garage-session-2/tree/main/.github/skills/filesystem-context
Command: npx skills add https://github.com/bthillerup/bens-garage-session-2 --skill filesystem-context-bthillerup

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill addresses the challenge of managing large amounts of context that exceed the agent's context window, enabling more efficient and scalable agent operations.

Core Features & Use Cases

  • Context Offloading: Persist large tool outputs or intermediate states to files to free up the context window.
  • Dynamic Context Discovery: Load relevant information from the filesystem on demand, rather than keeping everything in memory.
  • Agent Memory: Use files to maintain state across long-running tasks or between agent interactions.
  • Inter-Agent Communication: Facilitate information sharing between sub-agents through shared file workspaces.

Quick Start

Use the filesystem-context skill to save the output of the last tool execution to a file named 'tool_output.txt'.

Frequently Asked Questions about filesystem-context

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

FAQPage Schema
How do I manage agent context window bloat from large tool outputs?▼

Agent memory persistence uses the filesystem to maintain state across long-running tasks and interactions. By saving intermediate states to files, agents can dynamically load relevant context on demand rather than keeping everything in memory.

Can I use filesystem workspaces for inter-agent communication?▼

Dynamic context discovery loads relevant information from the filesystem on demand rather than keeping everything in memory. The agent retrieves persisted tool outputs and intermediate states from files only when needed during task execution.

What's the best way to persist agent state for long-running tasks?▼

To save tool output to a file, use the filesystem-context skill to persist the last tool execution output directly to a named file. This immediately frees up the context window while retaining the data for later dynamic loading.

Why does my agent lose context during extended multi-step operations?▼

Agents lose context during extended operations when intermediate states exceed the context window limits. Offloading data to the filesystem provides persistent agent memory, allowing state recovery and dynamic context discovery across long-running tasks.

Related Skills

Explore More Software Engineering Skills→
Tigha66Tigha66

filesystem-mcp

Enforce sandboxed filesystem access with directory permissions and audit trails.

Community
Advanced
samvanmesamvanme

filesystem-context

Offload large tool outputs and persistent context to files using Python and JSON/YAML.

Community
Basic
elihuvillarauselihuvillaraus
2

always-on-memory

Synchronize persistent coding session context across Claude, Gemini, and GitHub Copilot.

Community
Advanced
marinvchmarinvch

filesystem-context

Offload tool outputs and context to the filesystem for persistent agent workflows.

Community
Advanced