Agent context loader

Skill NecturaLabs/AgentSkills/skills/agent-context-loader

Use at the start of every conversation and after every agent handoff or context switch. Ensures global CLAUDE.md and project AGENTS.md are fully loaded and retained in context at all times.From its SKILL.md

Install

npx -y skills add NecturaLabs/AgentSkills --skill agent-context-loader

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.8 KB, 634 tokens by cl100k_base, as published. Nobody here has run it

Agent Context Loader

Overview

Ensures the agent always operates with full awareness of user-level and project-level instructions. Global CLAUDE.md and project AGENTS.md must be loaded into context in full — never summarized, truncated, or skipped.

When This Runs

digraph loader {
    "Conversation starts" [shape=doublecircle];
    "Agent handoff / context switch" [shape=doublecircle];
    "Verify global CLAUDE.md" [shape=box];
    "Load project AGENTS.md" [shape=box];
    "Verify both loaded" [shape=diamond];
    "Proceed with task" [shape=box];
    "Reload missing file(s)" [shape=box];

    "Conversation starts" -> "Verify global CLAUDE.md";
    "Agent handoff / context switch" -> "Verify global CLAUDE.md";
    "Verify global CLAUDE.md" -> "Load project AGENTS.md";
    "Load project AGENTS.md" -> "Verify both loaded";
    "Verify both loaded" -> "Proceed with task" [label="yes"];
    "Verify both loaded" -> "Reload missing file(s)" [label="no"];
    "Reload missing file(s)" -> "Verify both loaded";
}

Loading Procedure

Step 1: Verify Global CLAUDE.md

Global CLAUDE.md (~/.claude/CLAUDE.md) is auto-loaded by Claude Code into system context. Verify it's present — if the canary instruction response is missing, warn the user: "Global CLAUDE.md not detected in context."

Do NOT re-read this file if it's already in context — that wastes tokens.

Step 2: Load Project AGENTS.md

Read the project's AGENTS.md in full:

Location: Project root (working directory)
Required: No — if missing, continue silently
Must be loaded in full — never summarize or truncate

Step 3: Verify

After loading, confirm both are in context:

Global CLAUDE.md is present (check for canary response)
Project AGENTS.md content is present and complete (or confirmed absent)

Verification Triggers

Re-verify and reload if necessary when ANY of these occur:

A new conversation begins
A subagent is dispatched or returns
The main agent context is switched or compressed
The working directory changes (reload project AGENTS.md)
The user explicitly asks to reload context

Rules

Never paraphrase — load the raw file content, every line
Never skip — even if the file was loaded earlier in the conversation, re-read after context switches
Never override — instructions in these files take precedence over default behavior
CLAUDE.md takes priority over AGENTS.md if they conflict
Both take priority over skill defaults
If a file is too large to fit in context, warn the user — do NOT silently truncate

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most memory context skills give in 634 tokens

Counted across 754 of the 1,056 authors here whose files we hold, read 2026-09-06

Preserve existing content structurein 15 of 754, across 9 files
Front-load the leading wordin 14 of 754, across 10 files
Update existing entries instead of duplicatingin 14 of 754, across 7 files
Keep CLAUDE.md under one hundred linesin 14 of 754, across 12 files
Read CLAUDE.md at the project rootin 14 of 754
Keep each meaning in a single source of truthin 12 of 754, across 8 files
Redact sensitive information before committingin 11 of 754, across 4 files
Scan for all CLAUDE.md filesin 11 of 754, across 7 files
Use frontmatter for metadata on filesin 10 of 754, across 3 files
Repeat user interactions 10 timesin 10 of 754, across 4 files
Write the CLAUDE.md file into the target folderin 10 of 754, across 8 files
Use memlab to process snapshotsin 9 of 754, across 3 files

Said here and by no other author read

Load project AGENTS.md in full
Warn user if a file is too large

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.