black-box-supervisor

black-box-supervisor is a skill for Claude Code, Codex from X-School-Academy/skill-pilot. It costs 114 tokens per session (4,142 once invoked), scanned C, original, MIT.

A method for supervising coding agents from outside the project by giving them tasks in terminal windows and checking their results as a user would.

In plain words
What is it for?
It is for guiding agents through coding work, testing running applications, checking browser behavior and APIs, and requesting fixes.
Why use it?
It separates task direction and real-world checking from code editing, making it easier to find visible, API, log, and runtime problems.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: mentions Claude Code; mentions Codex; mentions Gemini CLI.

Good fit It is for guiding agents through coding work, testing running applications, checking browser behavior and APIs, and requesting fixes.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/x-school-academy/skill-pilot/black-box-supervisor
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.

Any agent
npx skills add X-School-Academy/skill-pilot --skill black-box-supervisor
Clone the repo
git clone --depth 1 https://github.com/X-School-Academy/skill-pilot

Made for: Claude Code, Codex.

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 black-box-supervisor

README.md
[![agentmods](https://agentmods.dev/badge/skills/x-school-academy/skill-pilot/black-box-supervisor/github.svg)](https://agentmods.dev/skills/x-school-academy/skill-pilot/black-box-supervisor)
Your own site
<a href="https://agentmods.dev/skills/x-school-academy/skill-pilot/black-box-supervisor"><img src="https://agentmods.dev/badge/skills/x-school-academy/skill-pilot/black-box-supervisor/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for black-box-supervisor

Your own site · 80×15
<a href="https://agentmods.dev/skills/x-school-academy/skill-pilot/black-box-supervisor"><img src="https://agentmods.dev/badge/skills/x-school-academy/skill-pilot/black-box-supervisor.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 114 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 4,142 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 2 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00114 $0.04142
Opus 5.5 $0.00046 $0.01657
Sonnet 5.5 $0.00023 $0.00828
Haiku 4.5 $0.00011 $0.00414

Measured 16d ago against content hash fcf1df9a6a9e, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade C, and why

black-box-supervisor scanned grade C with 2 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 16d 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.

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

rm -rf

Makes network callslowCapability

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

description: Use this skill when acting as a Black-Box Code Agent Supervisor. The supervisor uses coding agent CLIs such as Claude Code, Codex, OpenCode, Gemini CLI, or other AI coding agents through tmux terminal window
core/skills/system/black-box-supervisor/SKILL.md · 876 lines

How it starts

The opening of the file, as written. The whole thing — 876 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Black-Box Code Agent Supervisor

You are a Black-Box Code Agent Supervisor.

Your task is to use coding agent CLIs, such as:

  • Claude Code
  • Codex
  • OpenCode
  • Gemini CLI
  • other AI coding agents

through tmux terminal windows to finish a task or coding project.

You work like a real developer who uses AI coding agents to get work done.

You do not directly inspect source code.

You do not directly write code.

You do not directly write code-level tests.

Instead, you guide the coding agent, check the running result, find problems from the outside, and send clear instructions back to the coding agent.


1. Core idea

The coding agent is responsible for:

  • reading source code
  • writing code
  • editing files
  • fixing implementation issues
  • running tests
  • explaining technical changes

You are responsible for:

  • giving clear task instructions
  • checking the running result
  • checking the product as a non-technical customer
  • using the browser to find visible issues
  • using curl to find API issues
  • checking logs and terminal output
  • asking the coding agent to investigate and fix problems
  • asking the coding agent to verify its own work
  • deciding whether the result is acceptable

You are a black-box supervisor.

You judge the work by what the system does, not by reading how the code is written.


2. Hard rules

Never inspect source code directly

You must not directly read, search, or inspect implementation files.

Do not use commands like:

cat src/...
sed -n ... src/...
grep ... src/...
find . -name "*.ts"
find . -name "*.py"
git diff
git show
git log -p
git status --short

Do not inspect:

  • source files
  • frontend components
  • backend route handlers
  • controllers
  • services
  • database migration files
  • test files
  • package internals
  • generated build files
  • implementation diffs

If you need to know something about the code, ask the coding agent to inspect it and report back.


Never write code directly

You must not create, edit, patch, or rewrite project files.

Read the full file on GitHub · 876 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. 16d ago First seen · 876 lines · 0 tokens per session scan C fcf1df9a6a9e

Subscribe to this mod's changes

black-box-supervisor is a skill published in the GitHub repository X-School-Academy/skill-pilot (20 stars, last pushed 4mo ago), licensed MIT. It adds 114 tokens to every session and 4,142 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it C with 2 findings (recursive force delete, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-02.

Related

Other skills, from other repositories

debug-optimize-lcp

Guides debugging and optimizing Largest Contentful Paint (LCP) using Chrome DevTools MCP tools. Use this skill whenever the user asks about LCP performance, slow page loads, Core Web Vitals optimization, or wants to understand why their page's main content takes too long to appear. Also use when the user mentions…

ChromeDevTools/chrome-devtools-mcp · 99 tokens

systematic-debugging

Use when debugging a failing test, build error, or runtime issue that isn't immediately obvious. Guides a 4-phase root cause analysis instead of random fix attempts.

open-metadata/OpenMetadata · 37 tokens

log-error-digest

Analyze log files to troubleshoot errors, identify peak error periods, and produce error clustering, frequency statistics, and time distribution reports. Supports JSON, syslog, and Nginx formats with automatic detection. Use when a user uploads a .log file and asks to analyze errors, find patterns, debug issues, or…

zebbern/claude-code-guide · 71 tokens

deck-retro

Run a fully local agent-deck retrospective over the user's own transcripts, Recall index and logs. Use when a user wants to find recurring failures or take their own finding from investigation through synthetic reproduction, a user-filed issue, a test-first fix and contributor PR handoff. Also use for weekly usage…

asheshgoplani/agent-deck · 86 tokens

incident-response-incident-response

Coordinate production incident triage, mitigation, investigation, recovery, and follow-up. Use when responding to outages or other incidents that require structured operational coordination.

rmyndharis/antigravity-skills · 36 tokens

ci-fixer

CI failures - read error, minimal fix, verify.

sipyourdrink-ltd/bernstein · 15 tokens