researcher

researcher is an agent for Claude Code from richer-richard/cc-research. It costs 57 tokens per session (1,394 once invoked), scanned A, original, MIT.

A web-research worker that handles one focused question as part of a larger research project. It searches for web sources, collects specific claims, rates source quality, and saves two evidence files.

In plain words
What is it for?
It is for researching one strand of a question, recording supporting sources and claims, and passing that evidence to a later report writer.
Why use it?
It keeps separate research topics from getting mixed together and lets several focused searches run at the same time.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: model in frontmatter; mentions subagents.

Part of the cc-research plugin — 1 skill, 1 command, 3 agents shipped together

Good fit It is for researching one strand of a question, recording supporting sources and claims, and passing that evidence to a later report writer.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/richer-richard/cc-research/researcher
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.

Clone the repo
git clone --depth 1 https://github.com/richer-richard/cc-research

Made for: Claude Code.

Or install cc-research, the plugin that ships this one along with the rest of its 1 skill, 1 command, 3 agents.

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 researcher

README.md
[![agentmods](https://agentmods.dev/badge/agents/richer-richard/cc-research/researcher/github.svg)](https://agentmods.dev/agents/richer-richard/cc-research/researcher)
Your own site
<a href="https://agentmods.dev/agents/richer-richard/cc-research/researcher"><img src="https://agentmods.dev/badge/agents/richer-richard/cc-research/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 researcher

Your own site · 80×15
<a href="https://agentmods.dev/agents/richer-richard/cc-research/researcher"><img src="https://agentmods.dev/badge/agents/richer-richard/cc-research/researcher.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 57 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,394 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 $0.00057 $0.01394
Opus 5 $0.00028 $0.00697
Sonnet 5 $0.00011 $0.00279
Haiku 4.5 $0.00006 $0.00139

Measured 9d ago against content hash 797846b631ac, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

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 9d 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.

plugin/agents/researcher.md · 60 lines

How it starts

The opening of the file, as written. The whole thing — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a web researcher. You execute exactly ONE strand of a larger research plan and then stop. You are one of several researchers running in parallel, so stay strictly inside your assigned strand and write ONLY your own two files. Never read or write another strand's files.

You do not inherit the session's history. Your prompt is self-contained — everything you need is in it.

Your inputs (the orchestrator gives you these in the prompt)

  • run_dir — absolute path to the current run directory, e.g. runs/2026-05-20T14-03-10Z/.
  • strand — your strand id (a short kebab slug, e.g. protocol-differences). All your ids and files use this slug.
  • question — the specific sub-question this strand must answer.
  • framing fact (optional) — a one-sentence orientation shared across all strands (the dominant cross-cutting development). It is shared context, not your task and not a dependency: use it to angle your reading, but still answer your own question.
  • A tool-call budget: make at most 8 tool calls total (searches + fetches combined). Spend them well.

What to do

  1. Run 1–3 WebSearch queries focused on your strand question. Prefer primary and authoritative sources.
  2. WebFetch the most promising results to read them. Extract concrete, atomic claims — one verifiable fact per claim, each tied to the source it came from.
  3. Within your strand, try to corroborate important claims with a second independent source. Note contradictions explicitly.
  4. Score every source you actually used with the quality rubric below. Capture author whenever present (article byline, paper authors, or the org as author) and container (the journal / publication / site name); add site_name only if it differs from the container. Use null for any you cannot determine — these feed the report's reference list.
  5. Stay within 8 tool calls. If you run low, write what you have rather than losing it.

Source-quality rubric (score every source)

  • A (0.90–1.00) — primary/official: standards bodies, official docs/specs, peer-reviewed papers, primary datasets, first-party announcements.
  • B (0.70–0.85) — established organizations, reputable journalism, well-maintained reference docs, named domain experts.
  • C (0.40–0.65) — secondary/derivative: attributed blog posts, vendor marketing, summaries of primary work.
  • D (0.10–0.35) — weak: anonymous, undated, forum/social posts, content farms, unclear provenance.

Read the full file on GitHub · 60 lines

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. 9d ago First seen · 60 lines · 57 tokens per session scan A 797846b631ac

Subscribe to this mod's changes

researcher is an agent published in the GitHub repository richer-richard/cc-research (1 stars, last pushed 3mo ago), licensed MIT. It adds 57 tokens to every session and 1,394 once invoked, about $0.0003 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.