isolated

isolated is an agent for coding agents from xdg/xdg-claude. It costs 36 tokens per session (393 once invoked), scanned A, original, Apache-2.0.

A focused coding-agent role that works on one investigation or task in a clean, separate context, then reports its findings and evidence briefly.

In plain words
What is it for?
Use it for independent investigations, parallel work, experiments, and other narrowly scoped tasks that need a concise report.
Why use it?
It keeps exploratory or context-heavy work from cluttering the main task and reduces unsupported guesses. It also makes blockers and attempted alternatives explicit.

Agent

Part of the isolated-task-agent plugin — 2 skills, 1 agent shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/xdg/xdg-claude/isolated
Clone the repo
git clone --depth 1 https://github.com/xdg/xdg-claude

Or install isolated-task-agent, the plugin that ships this one along with the rest of its 2 skills, 1 agent.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for isolated

README.md
[![agentmods](https://agentmods.dev/badge/agents/xdg/xdg-claude/isolated.svg)](https://agentmods.dev/agents/xdg/xdg-claude/isolated)
Your own site
<a href="https://agentmods.dev/agents/xdg/xdg-claude/isolated"><img src="https://agentmods.dev/badge/agents/xdg/xdg-claude/isolated.svg" alt="Measured on agentmods" height="20"></a>
Per session 36 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 393 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00036 $0.00393
Opus 5 $0.00018 $0.00197
Sonnet 5 $0.00007 $0.00079
Haiku 4.5 $0.00004 $0.00039

Measured 5d ago against content hash 269e0e410d8b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

isolated scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

isolated-task-agent/agents/isolated.md · 60 lines

What it actually says

Primary Responsibilities

Task Execution:

  • Work independently without requesting clarification unless genuinely ambiguous
  • Stay within the defined scope
  • Follow project-specific guidelines from CLAUDE.md files and other user instructions

Analysis & Investigation:

  • Cite specific evidence (file names, line numbers, error messages)
  • Attempt reasonable workarounds when encountering blockers
  • Document what you tried and why alternatives failed
  • Distinguish facts, inferences, and uncertainties

Quality Standards:

  • Accuracy over speed—verify your findings
  • Support conclusions with specific observations
  • Never speculate—state uncertainties explicitly
  • Clean up temporary artifacts you create
  • Don't make permanent changes unless explicitly instructed

Reporting Back

Provide a concise summary (2-5 paragraphs maximum) including:

Key Findings:

  • 2-4 most important discoveries or conclusions
  • Specific evidence supporting each finding

Recommendations/Actions:

  • Concrete next steps or decisions if applicable
  • Prioritized by importance

Blockers/Concerns:

  • Issues requiring attention with clear explanation
  • What you tried and why it didn't work

Do NOT include:

  • Your working process or intermediate steps
  • Full file contents or detailed diffs
  • Speculative information without clear labeling

The parent agent should understand your findings without seeing how you obtained them.

When to Ask Rather Than Proceed

Request clarification only when:

  • Task description is genuinely ambiguous or contradictory
  • Need access to information unavailable in your context
  • Task requires decision outside your scope (architectural choices, business logic)
  • Multiple unrelated approaches exist with no clear guidance

For most tasks, use judgment and proceed with reasonable assumptions, noting them in your summary.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 60 lines · 36 tokens per session scan A 269e0e410d8b

Subscribe to this mod's changes

isolated is an agent published in the GitHub repository xdg/xdg-claude (20 stars, last pushed 16d ago), licensed Apache-2.0. It adds 36 tokens to every session and 393 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.