CONTEXTSCOUT_TEST_FINDINGS

CONTEXTSCOUT_TEST_FINDINGS is an agent for coding agents from darrenhinde/OpenAgentsControl. It costs 0 tokens per session (1,982 once invoked), scanned A, original, MIT.

A test report for checking whether coding agents use ContextScout to find relevant project files and background information. It records the tests created and the results of their first run.

In plain words
What is it for?
Reviewing ContextScout integration tests, comparing expected and actual agent behaviour, and deciding what discovery workflows need improvement.
Why use it?
It shows where the agents fail to discover context automatically or give enough explanation. This helps identify gaps in how they search unfamiliar parts of a project.

Agent

About the project

OpenAgents Control is a framework that adds plan-first, approval-based coding workflows to OpenCode, an AI coding framework. Developers and teams use it to make agents follow shared coding patterns, work across several programming languages, and validate changes through testing and review. The catalogue includes its agents, skills, commands, plugin, and hook.

darrenhinde/OpenAgentsControl · 4,823 stars · on GitHub

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/darrenhinde/openagentscontrol/contextscout_test_findings
Clone the repo
git clone --depth 1 https://github.com/darrenhinde/OpenAgentsControl

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 CONTEXTSCOUT_TEST_FINDINGS

README.md
[![agentmods](https://agentmods.dev/badge/agents/darrenhinde/openagentscontrol/contextscout_test_findings.svg)](https://agentmods.dev/agents/darrenhinde/openagentscontrol/contextscout_test_findings)
Your own site
<a href="https://agentmods.dev/agents/darrenhinde/openagentscontrol/contextscout_test_findings"><img src="https://agentmods.dev/badge/agents/darrenhinde/openagentscontrol/contextscout_test_findings.svg" alt="Measured on agentmods" height="20"></a>
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 1,982 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 $0.00000 $0.01982
Opus 5 $0.00000 $0.00991
Sonnet 5 $0.00000 $0.00396
Haiku 4.5 $0.00000 $0.00198

Measured 5d ago against content hash 6e2912de784d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

CONTEXTSCOUT_TEST_FINDINGS 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 5d 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.

evals/agents/CONTEXTSCOUT_TEST_FINDINGS.md · 241 lines

How it starts

The opening of the file, as written. The whole thing — 241 lines — stays where its author put it; the contents beside it link to each section on GitHub.

ContextScout Integration Test Findings

Date: 2026-01-09
Status: Tests Created, Initial Run Complete


Summary

We created comprehensive tests for ContextScout integration and ran initial tests. Here's what we learned:


Key Findings

1. OpenAgent Does NOT Use ContextScout Proactively ❌

Test: 04-implicit-discovery.yaml - "How does the registry system work?"

Expected: OpenAgent should delegate to ContextScout to discover registry context files

Actual: OpenAgent used grep and read directly to find information

  • Used grep to search for "registry"
  • Read registry.json directly
  • Used grep to search for "auto-detect"
  • Eventually found .opencode/context/openagents-repo/core-concepts/registry.md
  • Did NOT use the task tool to delegate to ContextScout

Conclusion: OpenAgent is NOT proactively using ContextScout for discovery. It's doing its own searching.


2. OpenAgent Provides Minimal Responses for Discovery Requests ⚠️

Test: 02-unknown-domain-discovery.yaml - "Explain how the eval framework works"

Expected: OpenAgent should use ContextScout and provide comprehensive answer

Actual: OpenAgent made only 1 tool call and provided minimal response

  • Very short execution time (14.5s)
  • Only 1 tool call (insufficient for discovery)
  • Did NOT delegate to ContextScout

Conclusion: OpenAgent may be treating discovery requests as conversational rather than requiring deep context loading.


3. ContextScout Cannot Be Tested in True Standalone Mode 🔧

Test: 01-code-standards-discovery.yaml via --subagent=contextscout

Expected: ContextScout runs as mode: primary and uses glob/read directly

Actual: Even in "standalone" mode, ContextScout is wrapped by a parent agent

  • Test runner forces mode: primary (confirmed in debug logs)
  • But ContextScout still tries to delegate to itself via task tool
  • Tool calls show: task → ContextScout (recursive!)
  • The test framework captures parent agent's tool calls, not nested subagent's

Read the full file on GitHub · 241 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. 5d ago First seen · 241 lines · 0 tokens per session scan A 6e2912de784d

Subscribe to this mod's changes

CONTEXTSCOUT_TEST_FINDINGS is an agent published in the GitHub repository darrenhinde/OpenAgentsControl (4,823 stars, last pushed 3d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,982 tokens. 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.

Related

Other agents, from other repositories

.NET-Notebook-Migration-Agent

Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.

microsoft/ai-agents-for-beginners · 33 tokens

ralph-loop-runner

Use this agent when you need to execute a Ralph orchestration loop end-to-end and verify its completion. This includes testing prompts against the Ralph system, validating that orchestration completes successfully, and capturing both results and any runtime issues. Examples:\n\n \nContext: User wants to test if a…

mikeyobrien/ralph-orchestrator · 286 tokens

chaos-monkey

You are the Chaos Monkey ("Kaos Maymunu") — a mutation-testing saboteur for the WrongStack fleet. Your job is to prove whether a test suite actually pins down the code it claims to cover, by deliberately breaking that code and watching which mutants survive.

WrongStack/WrongStack · 0 tokens

26-governance-ipo

You are the Company Secretary and Head of Investor Relations building the governance infrastructure that takes a solo founder's idea to a publicly listed large-cap company. You ensure the company is institutionally managed, investor-ready at every stage, and prepared for the most scrutinized event in a company's life…

ankitjha67/product-architect · 0 tokens

33-partnerships-bizdev

You are the Head of Business Development & Partnerships. You grow the business through other companies' products, customers, and distribution rather than building everything yourself. You decide what to build, buy, or partner; you structure the deals; and you run the partner lifecycle from recruit to revenue. You…

ankitjha67/product-architect · 0 tokens

45-corporate-development

You are the Head of Corporate Development. You own inorganic growth - the things the company buys, invests in, or sells rather than builds: acquisitions, minority investments, joint ventures, and divestitures. Where BD & Partnerships (Agent 33) owns contractual growth (deals where two companies stay separate and…

ankitjha67/product-architect · 0 tokens