Skip to content
Agentmart
Submit a tool

context-management

Manage AI context health — prevent context rot, extract knowledge before compaction, compose reference docs.

Skill by franciscsimon

IndexedMIT★ 1
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/franciscsimon-context-management

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Manage AI context health — prevent context rot, extract knowledge before compaction, compose reference docs. Use when sessions get long or context feels degraded.

Reviews

Write a review

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/context-management

What's inside

Skill name
context-management
Description
Manage AI context health — prevent context rot, extract knowledge before compaction, compose reference docs. Use when sessions get long or context feels degraded.
License
MIT
Instructions
39 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Context Management

## When to Use

- Session has been running for 15+ turns
- Agent starts forgetting earlier instructions
- Payload size exceeds 80KB (check with `/noise status`)
- Before compacting a session

## Workflow

1. **Check context health**
   - How many turns have elapsed?
   - Is the agent following ground rules from the start of the session?
   - Use `/antipatterns` to check for Perfect Recall Fallacy

2. **Extract knowledge before it's lost**
   - `/checkpoint save pre-compact` — save a knowledge checkpoint
   - Summarize key decisions made so far
   - Note any discoveries or gotchas

3. **Reduce noise**
   - `/noise quiet` — trim verbose tool output
   - `/zoom out` — get concise responses

4. **Compose only what's needed**
   - `/ref list` — see available reference docs
   - `/ref load <relevant-doc>` — load only what's needed now
   - `/compose add <knowledge-file>` — add relevant knowledge

5. **Compact if needed**
   - `/compact` — with knowledge already extracted

6. **Re-inject critical context**
   - `/reminders list` — verify reminders are still active
   - `/role <name>` — re-activate role if needed
   - Re-state the current task explicitly

From franciscsimon/harness, MIT.

More from franciscsimon/harness

Signals

Stars
1
Forks
1
Open issues
0
Last commit (this folder)
Mar 16, 2026 (7 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
Repository
github.com/franciscsimon/harness
First seen
Oct 24, 2026
Last checked
Oct 24, 2026
Source
GitHubgithub-graphql

More from franciscsimon

Cite this page

context-management skill. Agent Mart. https://agentmart.dev/skills/franciscsimon-context-management (last updated 2026-10-24)