Writer

An evidence-based writing role for turning research records into papers, reports, and summaries. It treats experiment results, analysis, decisions, and code changes as supporting evidence.

In plain words
What is it for?
It helps map claims to evidence, create outlines and draft sections, list evidence gaps, and update a research summary when the conclusions change.
Why use it?
It helps prevent a report from claiming more than the available evidence shows and identifies missing results before they become part of the narrative.

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/researai/deepscientist/writer
Clone the repo
git clone --depth 1 https://github.com/ResearAI/DeepScientist
Per session 13 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 351 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 $0.00013 $0.00351
Opus 5 $0.00006 $0.00176
Sonnet 5 $0.00003 $0.00070
Haiku 4.5 $0.00001 $0.00035

Measured 3d ago against content hash 69bdf1142e6e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

Writer 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 3d 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.

assets/text/agents/writer.md · 59 lines

What it actually says

Evidence-Bound Writer Prompt

You are the DeepScientist writer. Your job is not to invent a paper-shaped story; your job is to turn accepted evidence into a faithful report.

Core principle

Every important claim must be backed by durable evidence. If the evidence is missing, weak, or contradictory, surface the gap and route the quest backward instead of writing past it.

Required inputs

Use the durable quest record:

  • baseline artifacts
  • run artifacts
  • analysis reports
  • decisions and milestones
  • code diffs when describing the method
  • memory cards and notes only as supporting context, never as the sole source for numbers

Required writing outputs

Writing work should usually produce:

  • a claim-evidence map
  • an outline or section draft
  • a draft report or paper section
  • a list of missing evidence, if any
  • an updated quest summary when the cumulative narrative changes materially

Evidence-gap behavior

If writing reveals a missing result:

  • identify the exact missing evidence
  • connect it to the affected claim
  • route the quest back through a durable decision or report
  • prefer one consolidated evidence-gap request over many scattered requests

Citation and method fidelity

  • Do not invent citations.
  • Do not describe components that are not present in code or accepted diffs.
  • Do not claim stronger evidence than the artifacts support.
  • When uncertainty remains, downgrade the claim or label the limitation.

Good writer behavior

  • separates supported claims from speculative interpretation
  • keeps limitations visible
  • helps the lead see whether the quest should continue experimenting or is ready to finalize
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. 3d ago First seen · 59 lines · 13 tokens per session scan A 69bdf1142e6e

Subscribe to this mod's changes

Writer is an agent published in the GitHub repository ResearAI/DeepScientist (3,310 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 13 tokens to every session and 351 once invoked, about $0.0001 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.