scout

scout is a command for coding agents from arslan70/haytham. It costs 33 tokens per session (1,669 once invoked), scanned A, original, MIT.

An automated daily process for finding and investigating startup ideas from a collected set of problems. It runs without waiting for a person to answer questions.

In plain words
What is it for?
Use it to check a harvest of collected problem data, screen candidates, choose at most one for deeper research, and save the run's status and report.
Why use it?
It removes repetitive screening and selection work from daily idea research. It also records failures and can produce a report even when no idea is selected.

Command

Part of the haytham plugin — 7 skills, 7 commands, 12 agents, 2 hooks shipped together

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 commands/arslan70/haytham/scout
Clone the repo
git clone --depth 1 https://github.com/arslan70/haytham

Or install haytham, the plugin that ships this one along with the rest of its 7 skills, 7 commands, 12 agents, 2 hooks.

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 scout

README.md
[![agentmods](https://agentmods.dev/badge/commands/arslan70/haytham/scout.svg)](https://agentmods.dev/commands/arslan70/haytham/scout)
Your own site
<a href="https://agentmods.dev/commands/arslan70/haytham/scout"><img src="https://agentmods.dev/badge/commands/arslan70/haytham/scout.svg" alt="Measured on agentmods" height="20"></a>
Per session 33 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,669 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 $0.00033 $0.01669
Opus 5 $0.00016 $0.00834
Sonnet 5 $0.00007 $0.00334
Haiku 4.5 $0.00003 $0.00167

Measured 4d ago against content hash f84bab44b4b7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

scout 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 4d 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.

commands/scout.md · 90 lines

How it starts

The opening of the file, as written. The whole thing — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Haytham: Daily Scout (autonomous)

You are running the autonomous scout pipeline headless. There is no human to ask; never stop for clarification. Deterministic decisions (score floor, dedup, winner selection) belong to scripts — you orchestrate, you do not decide.

IMPORTANT: Always read agent output from files, not from conversation history.

Hard rules

  1. Never restart or re-run the pipeline because a verdict is PIVOT or NO-GO. One pass per day; the verdict is the result.
  2. Maximum ONE deep dive per run.
  3. Never edit the ledger or selection files — scout_select.py owns them.
  4. If a stage fails or produces nothing, stop after recording what happened; the report script ships an honest-zero report from whatever artifacts exist. Shipping nothing is the only failure mode.

Setup

RUN_DIR = $ARGUMENTS (a path relative to the working directory).

  1. Verify RUN_DIR/harvest/telemetry.json exists. If not, write RUN_DIR/scout-status.json with {"failed_at": "setup", "reason": "no harvest"} and stop.
  2. Read PERSONA from config/persona.yaml if it exists; otherwise use: "Solo technical founder building with AI tooling; wants ideas testable as a 1-2 week solo MVP; avoids heavy ops, sales-led GTM, regulated domains."
  3. Call TodoWrite once with the steps below.

Step 1 — Extract

Replay guard: if RUN_DIR/candidates/candidates.json already exists and parses with a non-empty candidates list, skip this step (replay/sampling run) and go to Step 2.

Launch the idea-scout agent. Tell it the RUN_DIR and give it PERSONA verbatim. It writes RUN_DIR/candidates/candidates.json.

Verify the file exists and parses. If it is missing or candidates is empty, write RUN_DIR/scout-status.json {"failed_at": "extract", "reason": "<what happened>"} and stop.

Step 2 — Screen

Replay guard: if RUN_DIR/screen/screening.json already exists and parses with a non-empty scorecards list, skip this step and go to Step 3.

Launch the feasibility-screener agent. Tell it the RUN_DIR and give it PERSONA verbatim. It writes RUN_DIR/screen/screening.json and screening.md.

Read the full file on GitHub · 90 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. 4d ago First seen · 90 lines · 33 tokens per session scan A f84bab44b4b7

Subscribe to this mod's changes

scout is a command published in the GitHub repository arslan70/haytham (13 stars, last pushed 1mo ago), licensed MIT. It adds 33 tokens to every session and 1,669 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 2026-08-30.