skills: Agent for Claude Code

.claude/agents/token-caching-researcher.md

token-caching-researcher is an agent for Claude Code from JosephSanjaya/skills. It costs 40 tokens per session (246 once invoked), scanned A, original, MIT.

A technical research agent that examines local files for information about token caching, subprocess isolation, and cost control in Claude Code.

In plain words
What is it for?
Finding caching rules, subprocess settings, persistent JSON-stream details, tool-definition requirements, and related command-line options in a local codebase.
Why use it?
It gathers scattered implementation details into one concise report without changing files. Token caching means reusing processed prompt content to reduce repeated work; a subprocess is a separate running command or program.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: model in frontmatter; mentions Claude Code.

This is JosephSanjaya/skills's own configuration. It tells Claude Code how to work on skills itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything skills configures →

Reuse

Borrowing it

Nothing to install: this file belongs to JosephSanjaya/skills. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/JosephSanjaya/skills/main/.claude/agents/token-caching-researcher.md
Clone the repo
git clone --depth 1 https://github.com/JosephSanjaya/skills

Made for: Claude Code.

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 token-caching-researcher

README.md
[![agentmods](https://agentmods.dev/badge/agents/josephsanjaya/skills/token-caching-researcher/github.svg)](https://agentmods.dev/agents/josephsanjaya/skills/token-caching-researcher)
Your own site
<a href="https://agentmods.dev/agents/josephsanjaya/skills/token-caching-researcher"><img src="https://agentmods.dev/badge/agents/josephsanjaya/skills/token-caching-researcher/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for token-caching-researcher

Your own site · 80×15
<a href="https://agentmods.dev/agents/josephsanjaya/skills/token-caching-researcher"><img src="https://agentmods.dev/badge/agents/josephsanjaya/skills/token-caching-researcher.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 40 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 246 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 $0.00040 $0.00246
Opus 5 $0.00020 $0.00123
Sonnet 5 $0.00008 $0.00049
Haiku 4.5 $0.00004 $0.00025

Measured 11d ago against content hash 96cd13ca5dec, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

token-caching-researcher 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 11d 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.

.claude/agents/token-caching-researcher.md · 27 lines

What it actually says

Role

You are a specialized technical researcher. Your task is to analyze the local codebase and customizations directory to gather technical details about prompt caching and subprocess isolation in Claude Code.

Scope

  • Search for files containing information on prompt caching, subprocess wrappers, settings, or CLI flags.
  • Extract the 4-layer subprocess isolation rules and CLI configurations.
  • Extract details on persistent JSON streams and tool definition stability.
  • Summarize these findings into a concise, high-signal technical report.

Task Boundaries

  • Do NOT edit or write any files.
  • Do NOT run shell commands.

Output Format

Return a structured Markdown summary covering:

  1. 4-Layer Subprocess Isolation (exact steps and flags)
  2. Tool Definition Caching (alphabetical sorting, locked tools, stubs)
  3. Dynamic Loop Caching (cache_control placement, system-reminders)
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. 11d ago First seen · 27 lines · 40 tokens per session scan A 96cd13ca5dec

Subscribe to this mod's changes

token-caching-researcher is an agent published in the GitHub repository JosephSanjaya/skills (5 stars, last pushed 4d ago), licensed MIT. It adds 40 tokens to every session and 246 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 2026-08-31.