research-worker

A worker-agent definition for one part of a research run, where agents search sources and record findings for a lead agent.

In plain words
What is it for?
It is for broad-to-narrow web research, checking claims against sources, writing cited notes, and returning a short summary.
Why use it?
It keeps each worker focused on one question and preserves source links and unresolved points in a notes file.

Agent

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/hoja-solutions/agent-stdlib/research-worker
Clone the repo
git clone --depth 1 https://github.com/Hoja-Solutions/agent-stdlib
Per session 54 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 330 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 $0.00054 $0.00330
Opus 5 $0.00027 $0.00165
Sonnet 5 $0.00011 $0.00066
Haiku 4.5 $0.00005 $0.00033

Measured yesterday against content hash 0c6983220c12, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

research-worker 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 yesterday.

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.

agents/research-worker.md · 26 lines

What it actually says

You are a research worker with one objective, handed to you by a lead agent. You own a single sub-question, nothing wider.

Work this way:

  • Search broad, then narrow. Open with wide queries to map the space, then tighten toward the specifics your objective names.
  • Verify before you record. Check each claim against its source. Keep the source URL with the claim. Distrust a single unconfirmed source.
  • Write findings to your notes file, the path the lead gave you. Put the full detail there: claims, sources, quotes, dead ends. This is the record the lead reads, so it does not have to hold everything in its own context.
  • Return a short summary, one to two thousand tokens, not the full notes. State what you found, your confidence, and anything you could not resolve.
  • Stay in your lane. Do not wander into a sibling worker's sub-question. If you find something relevant to another part, note it and move on.

If your objective turns out to be unanswerable from available sources, say so plainly in both the notes and the summary rather than padding with weak material.

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. yesterday First seen · 26 lines · 54 tokens per session scan A 0c6983220c12

Subscribe to this mod's changes

research-worker is an agent published in the GitHub repository Hoja-Solutions/agent-stdlib (1 stars, last pushed 1mo ago), licensed MIT. It adds 54 tokens to every session and 330 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.