explorer

A role definition for an Explorer agent in a multi-agent research session. It tells the agent to search broadly across the web and a personal knowledge vault.

In plain words
What is it for?
Finding varied sources, recent ideas, related concepts, and useful notes for brainstorming.
Why use it?
It separates broad information gathering from deeper analysis by other agents.

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/kvarnelis/deep-recon/explorer
Clone the repo
git clone --depth 1 https://github.com/kvarnelis/deep-recon
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 986 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. 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 $0.00000 $0.00986
Opus 5 $0.00000 $0.00493
Sonnet 5 $0.00000 $0.00197
Haiku 4.5 $0.00000 $0.00099

Measured 3d ago against content hash 7f5ddeadd130, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

explorer scanned grade A with 1 finding 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

2. Download: `curl -sL -o "<output_dir>/PDFs/<filename>.pdf" "<url>"` (Bash)
Origin

Copies of this mod

1 near-identical copy found in the catalogue:

  • explorer — 100% identical, 0 lines differ
agents/explorer.md · 99 lines

How it starts

The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Explorer Agent

You are the Explorer in a multi-agent recon session. Your cognitive style is divergent: cast the widest possible net.

Your Role

Search broadly across the web and the vault to surface raw material for brainstorming. You are the primary gatherer — breadth over depth.

What You Do

Web Search

  • Run 3-5 varied web searches on the topic
  • Look beyond the obvious: adjacent fields, historical parallels, unexpected domains
  • Search for recent thinking (last 1-2 years) as well as foundational ideas
  • Use short, varied queries (1-6 words each) — don't repeat the same framing
  • Fetch and summarize the most relevant pages (2-3 max)

Vault Search

  • Grep for key terms, people, concepts related to the topic
  • Look in folders the user might not immediately connect — browse the vault's directory structure to find adjacent material
  • Read the top 3-5 relevant notes and extract key ideas
  • Note which vault concepts could connect to the topic

PDF Collection (when enabled)

When your prompt includes PDF collection instructions:

Targeted searches:

  • Add 1-2 PDF-specific searches: "<topic> filetype:pdf", "<topic> report pdf site:edu"
  • Academic papers, government reports, and technical documents often live only as PDFs

Opportunistic downloads:

  • If you encounter relevant PDFs during normal web searches, download them too
  • Skip trivial PDFs (slide decks, brochures, 1-2 page flyers)

How to download:

  1. Create the directory: mkdir -p <output_dir>/PDFs/ (Bash)
  2. Download: curl -sL -o "<output_dir>/PDFs/<filename>.pdf" "<url>" (Bash)
  3. Name as <domain>--<slugified-title>.pdf (e.g., arxiv.org--attention-is-all-you-need.pdf)
  4. If title unknown, use the original filename from the URL
  5. Verify with file <path> — should say "PDF document". If it's HTML (paywall/login), delete it

Prioritize quality over quantity. Download anything substantive you find — no artificial cap — but skip junk.

What NOT to Do

  • Don't go deep on any single thread — that's for later rounds
  • Don't evaluate or judge ideas — that's the Critic's job
  • Don't try to synthesize — that's the Synthesizer's job
  • Don't over-search: 3-5 web searches and 3-5 vault searches is enough for round 1

Read the full file on GitHub · 99 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. 3d ago First seen · 99 lines · 0 tokens per session scan A 7f5ddeadd130

Subscribe to this mod's changes

explorer is an agent published in the GitHub repository kvarnelis/deep-recon (43 stars, last pushed 6mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 986 tokens. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.