skills.rest
cbrostrom
cbrostromCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

scout

Evaluates external tools and ideas against an existing agent setup with adopt, borrow, or reject verdicts.

Updated Nov 25, 2024
Search Tags:#agent setup#overlap analysis#technology scouting#tool evaluation#decision workflow#primary source research#adopt borrow reject
One-click install
npx skills add https://github.com/cbrostrom/dotfiles --skill scout-cbrostrom
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: scout
Source: https://github.com/cbrostrom/dotfiles/tree/main/stow/agents/.agents/skills/scout
Command: npx skills add https://github.com/cbrostrom/dotfiles --skill scout-cbrostrom

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? It prevents impulsive adoption of new tools, frameworks, and workflows by forcing primary-source research, overlap analysis against the existing agent setup, and a clear verdict before any change is made. ## Core Features & Use Cases - Structured evaluation workflow: Checks a rejection log, frames the candidate's claim, researches primary sources, maps fit across setup layers (memory, agent behavior, harnesses, orchestration, efficiency, operations), and delivers a verdict. - Hard verdict contract: Every evaluation ends in exactly one of adopt, borrow, or reject, with a ranked list of borrowable ideas or a bounded reversible experiment. - Append-only rejection log: Records each rejection with claim, evidence, and revisit conditions in a Higgins vault file to avoid re-evaluating dismissed ideas. - Use Case: When you discover a new agent memory framework, run scout to compare it against your Higgins vault and existing retrieval stack, then receive a borrow verdict listing only the provenance and staleness ideas worth extracting. ## Quick Start Ask the agent to scout a specific project or idea, for example: scout the OKF Agent Memory project and tell me whether it is worth adopting for my setup.

Frequently Asked Questions about scout

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I evaluate a new tool before adding it to my agent setup?▼

Run the scout workflow on the candidate: it checks prior rejections, reads primary sources, maps overlap against your setup layers, and returns an adopt, borrow, or reject verdict. No changes are made without your explicit approval.

What does the scout skill do with rejected ideas?▼

Rejected ideas are appended to a scout-rejections log in the Higgins vault with the claim, verdict reason, evidence checked, and a condition that would trigger reconsideration. The log is append-only so dismissed ideas are not re-evaluated without new evidence.

Can scout install or modify my setup automatically?▼

No. Scout never adopts or edits the setup on its own. For an adopt verdict it presents the files, change intent, break risk, verification, and rollback plan, then waits for explicit approval before acting.

What is the difference between adopt, borrow, and reject verdicts?▼

Adopt means the candidate fills a demonstrated gap and a minimal reversible pilot is proposed. Borrow means only specific ideas are extracted as patches, tests, or knowledge deposits. Reject means the candidate does not beat the current setup and the rejection is logged.

When should I not use a structured tool evaluation workflow?▼

Skip it for trivial, easily reversible experiments where the cost of trying is lower than the cost of analysis. The workflow is designed for candidates that touch multiple setup layers or add tools, daemons, or recurring jobs.

Related Skills

Explore More Software Engineering Skills→
walsh-investmentwalsh-investment

learning-test-theater-patterns

Mine repository artifacts to surface candidate test theater patterns for review.

Official
Intermediate
evolution-foundationevolution-foundation
520

ops-change-request

Generates approval-ready change requests with impact analysis, risk evaluation, and rollback plans.

Official
Intermediate
AGMO-IncAGMO-Inc
4

benchmark

Execute benchmark commands and store latency and throughput metrics in Obsidian vaults.

Official
Advanced
skyllwtskyllwt
1.7k

exp-pilot-eval

Evaluates pilot experiment results and updates idea pages with pass, fail, or inconclusive verdicts.

Community
Intermediate