Agent skill

Shadow Agent

A skill your agent uses when you need a background monitoring agent that watches the fleet, detects anomalies, and can intervene on command.

MITAuto-check passed

Install Shadow Agent

skills CLI

$ npx skills add WrongStack/WrongStack --skill shadow-agent -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI

$ gh skill install WrongStack/WrongStack shadow-agent --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy

$ git clone --depth 1 https://github.com/WrongStack/WrongStack.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.wrongstack/skills/shadow-agent .claude/skills/shadow-agent && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Paste into Claude Code

Install the "shadow-agent" agent skill from https://github.com/WrongStack/WrongStack/tree/main/.wrongstack/skills/shadow-agent into .claude/skills/shadow-agent/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "shadow-agent", then confirm the skill loads.

Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

$skill-installer

$skill-installer install https://github.com/WrongStack/WrongStack/tree/main/.wrongstack/skills/shadow-agent

Type this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.

skills CLI

$ npx skills add WrongStack/WrongStack --skill shadow-agent -a codex

Project install goes to .agents/skills/; add -g for ~/.codex/skills/.

GitHub CLI

$ gh skill install WrongStack/WrongStack shadow-agent --agent codex

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/WrongStack/WrongStack.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.wrongstack/skills/shadow-agent .agents/skills/shadow-agent && rm -rf skills-src

Use ~/.agents/skills/ instead of .agents/skills for a personal install.

Codex skills documentation · loads skills from .agents/skills/

Paste into Codex

Install the "shadow-agent" agent skill from https://github.com/WrongStack/WrongStack/tree/main/.wrongstack/skills/shadow-agent into .agents/skills/shadow-agent/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "shadow-agent", then confirm the skill loads.

Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add WrongStack/WrongStack --skill shadow-agent -a cursor

Project install goes to .agents/skills/; add -g for ~/.cursor/skills/.

GitHub CLI

$ gh skill install WrongStack/WrongStack shadow-agent --agent cursor

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/WrongStack/WrongStack.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.wrongstack/skills/shadow-agent .cursor/skills/shadow-agent && rm -rf skills-src

Use ~/.cursor/skills/ instead of .cursor/skills for a personal install.

Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/

Paste into Cursor

Install the "shadow-agent" agent skill from https://github.com/WrongStack/WrongStack/tree/main/.wrongstack/skills/shadow-agent into .cursor/skills/shadow-agent/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "shadow-agent", then confirm the skill loads.

Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

gemini skills install

$ gemini skills install https://github.com/WrongStack/WrongStack.git --path .wrongstack/skills/shadow-agent

--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.

skills CLI

$ npx skills add WrongStack/WrongStack --skill shadow-agent -a gemini-cli

Project install goes to .agents/skills/; add -g for ~/.gemini/skills/.

GitHub CLI

$ gh skill install WrongStack/WrongStack shadow-agent --agent gemini-cli

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/WrongStack/WrongStack.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.wrongstack/skills/shadow-agent .gemini/skills/shadow-agent && rm -rf skills-src

Use ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.

Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/

Paste into Gemini CLI

Install the "shadow-agent" agent skill from https://github.com/WrongStack/WrongStack/tree/main/.wrongstack/skills/shadow-agent into .gemini/skills/shadow-agent/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "shadow-agent", then confirm the skill loads.

Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

GitHub CLI

$ gh skill install WrongStack/WrongStack shadow-agent

Installs for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).

skills CLI

$ npx skills add WrongStack/WrongStack --skill shadow-agent -a github-copilot

Project install goes to .agents/skills/; add -g for ~/.copilot/skills/.

Manual copy

$ git clone --depth 1 https://github.com/WrongStack/WrongStack.git skills-src && mkdir -p .github/skills && cp -r skills-src/.wrongstack/skills/shadow-agent .github/skills/shadow-agent && rm -rf skills-src

Use ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.

GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/

Paste into GitHub Copilot

Install the "shadow-agent" agent skill from https://github.com/WrongStack/WrongStack/tree/main/.wrongstack/skills/shadow-agent into .github/skills/shadow-agent/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "shadow-agent", then confirm the skill loads.

GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add WrongStack/WrongStack --skill shadow-agent -a opencode

OpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.

GitHub CLI

$ gh skill install WrongStack/WrongStack shadow-agent --agent opencode

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/WrongStack/WrongStack.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.wrongstack/skills/shadow-agent .opencode/skills/shadow-agent && rm -rf skills-src

Use ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.

OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/

Paste into OpenCode

Install the "shadow-agent" agent skill from https://github.com/WrongStack/WrongStack/tree/main/.wrongstack/skills/shadow-agent into .opencode/skills/shadow-agent/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "shadow-agent", then confirm the skill loads.

OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

Facts

Skill name: shadow-agent
GitHub stars: 370
Token cost: ~2.3k tokens
SKILL.md length: 524 words
Files: 1
Skills in repo: 38
Repo updated: Oct 7, 2026
First seen: Oct 7, 2026
Licence: MIT

At a glance

A skill your agent uses when you need a background monitoring agent that watches the fleet, detects anomalies, and can intervene on command.

Works in 6 steps: Fleet Monitoring (every 30s by default) → Mailbox Surveillance → Spike Detection → …
You need a background monitoring agent that watches the fleet
SKILL.md covers Overview, Architecture, Core Responsibilities and Data Structures, plus 5 more sections
Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Shadow Agent is an agent skill from WrongStack/WrongStack. Use this skill when you need a background monitoring agent that watches the fleet, detects anomalies, and can intervene on command. Triggers: user says "shadow", "monitoring agent", "fleet watcher", "hoop command", "spike detection".

Its SKILL.md is about 2.3k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The repository describes itself as: An AI coding agent that reads your code, edits files, runs commands, and reasons through bugs — across a terminal REPL, a full-screen TUI, and a browser UI, while you keep your… The licence is MIT.

When your agent uses it

You need a background monitoring agent that watches the fleet
Detects anomalies
Can intervene on command

Example prompts

“shadow”
“monitoring agent”
“fleet watcher”
“/shadow-agent”

Workflow steps

6 steps, taken from the step headings in SKILL.md.

Fleet Monitoring (every 30s by default)
Mailbox Surveillance
Spike Detection
Deterministic Scheduling
LLM Analysis
Intervention Commands

What it can do on your machine

Read from SKILL.md and the folder at commit 57f6018. It shows what the files ask for, not the result of running them.

Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
No scripts in the folder and no shell commands in SKILL.md (its code samples are typescript).
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Shadow Agent loads about 2.3k tokens when it runs. Until then it costs about 62 tokens; SKILL.md has 524 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it: ~62
When it runs · the whole SKILL.md, loaded when a task matches: ~2.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from WrongStack/WrongStack at commit 57f6018, republished under its MIT licence (© WrongStack). 524 words, ~2,255 tokens.

Download SKILL.mdSave it as .claude/skills/shadow-agent/SKILL.md (or your agent's skills folder).

name: shadow-agent
description: Use this skill when you need a background monitoring agent that watches the fleet, detects anomalies, and can intervene on command. Triggers: user says "shadow", "monitoring agent", "fleet watcher", "hoop command", "spike detection".
version: 1.0.0
required-capabilities: coordination.mailbox
required-tools: cron_schedule, mailbox, terminate_subagent
optional-capabilities: fleet.delegate, work.plan

Shadow Agent — Fleet Monitoring & Intervention

Overview

A deterministic background agent that monitors all agents in the fleet via scheduled heartbeat checks. Shadow Agents run silently, observe everything, and intervene only when explicitly commanded or when critical anomalies are detected.

Architecture

┌─────────────────────────────────────────────────────────────────┐
│                     Shadow Agent                                 │
│  ┌──────────────┐  ┌──────────────┐  ┌──────────────────────┐ │
│  │ Heartbeat    │  │ FleetBus     │  │ Mailbox Monitor      │ │
│  │ Scheduler    │  │ Subscriber   │  │ (read/query/broadcast)│ │
│  │ (cron)       │  │             │  │                      │ │
│  └──────┬───────┘  └──────┬───────┘  └──────────┬───────────┘ │
│         │                  │                     │              │
│         └──────────────────┼─────────────────────┘              │
│                            ▼                                    │
│                   ┌─────────────────┐                          │
│                   │ LLM Analyzer    │ ← model: configured-model  │
│                   │ (pattern detect)│   setmodel to change       │
│                   └────────┬────────┘                          │
│                            │                                   │
│         ┌─────────────────┼─────────────────┐                │
│         ▼                 ▼                 ▼                │
│  ┌────────────┐   ┌──────────────┐   ┌──────────────┐        │
│  │ Intervention│   │ Report       │   │ Spike        │        │
│  │ Handler    │   │ Generator    │   │ Detector     │        │
│  │ (hoop/etc) │   │ (status)    │   │ (burst tasks)│        │
│  └────────────┘   └──────────────┘   └──────────────┘        │
└─────────────────────────────────────────────────────────────────┘
                            │
                            ▼
              ┌─────────────────────────────┐
              │ FleetBus + Mailbox + Cron    │
              │ (cross-terminal visibility) │
              └─────────────────────────────┘

Core Responsibilities

1. Fleet Monitoring (every 30s by default)

Subscribe to FleetBus for all subagent lifecycle events
Call fleet (action: status) + fleet (action: health) on each heartbeat
Track what each agent is currently doing (via task descriptions)
Detect stuck agents (>5 min no events), idle agents, crashed agents

2. Mailbox Surveillance

Monitor all mailbox messages (broadcast + direct)
Track message patterns: who is talking to whom, frequency
Detect orphan tasks (assign without result), stale asks
Cross-session awareness: see agents from other terminals

3. Spike Detection

Track task duration per agent type
Detect instant start/stop patterns (spike tasks)
Flag agents that spawn and die within seconds
Report efficiency anomalies

4. Deterministic Scheduling

Uses cron_schedule for fixed-interval heartbeats
Configurable interval: shadow_interval_ms (default 30000)
No randomness — same state always produces same actions
Clean startup/shutdown via cron cancel on teardown

5. LLM Analysis

Uses the configured model without a built-in model default
setmodel <model-id> changes the analysis model
Prompt-engineered for pattern detection
Minimal token usage — snapshot analysis only

6. Intervention Commands

Commands received via mailbox or direct invocation:

Command	Action
`hoop <subagentId>`	Immediate terminate via `terminate_subagent`
`hoop all`	Terminate all running subagents
`shadow status`	Report current fleet snapshot
`shadow mute`	Pause heartbeat monitoring
`shadow resume`	Resume heartbeat monitoring
`shadow interval <ms>`	Change heartbeat interval
`shadow intervene <task>`	Assign a custom intervention task

Data Structures

ShadowState (persisted)

typescript

interface ShadowState {
  enabled: boolean;
  intervalMs: number;
  model: string;
  startTime: string; // ISO8601
  lastHeartbeat: string; // ISO8601
  knownAgents: Map<agentId, AgentSnapshot>;
  spikeHistory: SpikeEvent[];
  anomalyLog: Anomaly[];
}

AgentSnapshot

typescript

interface AgentSnapshot {
  agentId: string;
  sessionId: string;
  name: string;
  role: string;
  status: 'running' | 'idle' | 'stopped' | 'unknown';
  currentTask?: string;
  lastSeen: string;
  eventCount: number;
  startedAt: string;
}

SpikeEvent

typescript

interface SpikeEvent {
  agentId: string;
  spawnedAt: string;
  terminatedAt: string;
  durationMs: number;
  task?: string;
  reason?: string; // 'completed' | 'error' | 'killed' | 'timeout'
}

Anomaly

typescript

interface Anomaly {
  id: string;
  type: 'stuck_agent' | 'spike_task' | 'mailbox_loop' | 'budget_exhausted';
  severity: 'low' | 'medium' | 'high' | 'critical';
  agentId?: string;
  description: string;
  detectedAt: string;
  resolvedAt?: string;
}

Output Format

Status Report

## Shadow Agent Status — <timestamp>

**Fleet**: <N> agents tracked | <M> running | <K> idle | <L> stopped
**Heartbeat**: every <X>s | Last: <timestamp>
**Model**: <model-id>

### Active Agents
| Agent | Session | Role | Status | Task | Last Seen |
|-------|---------|------|--------|------|-----------|
| ... | ... | ... | ... | ... | ... |

### Recent Anomalies
- [HIGH] agent-xyz stuck for 5m — no events detected
- [MED] Spike: agent-abc ran for 3s then died

### Mailbox Summary
- Unread: <N> messages | Broadcasts: <M>
- Pending asks: <K> (oldest: <agent> <age>)

Intervention Log

## Shadow Intervention Log

| Time | Command | Target | Result |
|------|---------|--------|--------|
| 10:23:01 | hoop | subagent-xyz | terminated |
| 10:25:33 | hoop | all | 3 agents terminated |

Workflow

Startup

Read persisted ShadowState (if exists)
Subscribe to FleetBus (onAny handler for all events)
Schedule heartbeat cron job at configured interval
Send broadcast: shadow:started { intervalMs, model, startTime }

Show full SKILL.md (222 more words)Show less

Heartbeat (every intervalMs)

Call fleet (action: status) + fleet (action: health)
Compare against known agents — detect changes
For each new agent: record snapshot, log "agent joined"
For each missing agent: mark unknown, log "agent left"
Analyze for anomalies (stuck, spike, loop)
If anomaly detected AND auto_intervene enabled: act
If anomaly detected AND auto_intervene disabled: report only
Update ShadowState.lastHeartbeat

FleetBus Events

subagent.started  → record agent, log
subagent.stopped   → record termination, check for spike
subagent.error     → log error, flag anomaly
tool.executed      → increment eventCount for agent
task.assigned      → update currentTask for agent
task.completed     → clear currentTask, log duration

Mailbox Monitoring

On incoming message:
  - If type=control and body starts with "hoop": execute intervention
  - If type=ask with subject "shadow status": send full report
  - If type=broadcast: track in mailbox summary
  - If type=assign without result within 5min: flag orphan

Intervention (hoop)

Parse target (single agent, "all", or pattern)
Call terminate_subagent(target) for each
Log intervention with timestamp
Send result to mailbox (to=sender)
If target="all": also cancel all pending cron jobs

Configuration

Config Key	Type	Default	Description
`shadow_interval_ms`	number	30000	Heartbeat interval in ms
`shadow_model`	string	session model	Analysis model
`shadow_auto_intervene`	boolean	false	Auto-act on anomalies
`shadow_stuck_threshold_ms`	number	300000	5min = stuck agent
`shadow_spike_threshold_ms`	number	5000	5s = spike task

Skills in scope

fleet-management — uses fleet (action: status), fleet (action: health), terminate_subagent
mailbox — uses mailbox tools for monitoring and intervention
cron — uses cron_schedule for deterministic heartbeats
multi-agent — for spawning analysis subagents when needed
observability — for structured logging of anomalies

Anti-Patterns

No randomness: Deterministic — same input → same output
No blocking: Heartbeats are async, never block agent loop
No loud logging: Uses DEBUG level unless anomaly
No direct tool calls: Uses FleetBus subscription, not polling
Minimal footprint: Shadow state is small, persisted efficiently

Files

Just SKILL.md in .wrongstack/skills/shadow-agent of WrongStack/WrongStack.

Open the folder on GitHubat commit 57f6018

Compare with similar skills

Shadow Agent next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Shadow Agent compared with similar skills
Skill	Stars	Used in	Tokens	Auto-check	Licence	Repo updated
Shadow Agent this skillWrongStack/WrongStack	370	—	~2.3k	Automated safety check: Pass	MIT	yesterday
Python Background Jobswshobson/agents	40k	—	~1.8k	Automated safety check: Pass	MIT	3 days ago
Backgroundjeremylongshore/tons-of-skills-marketplace	2.8k	—	~2.2k	Automated safety check: Pass	MIT	today
Canary Watchaffaan-m/ECC	275k	1 repos	~770	Automated safety check: Pass	MIT	3 days ago
Canary Watchaffaan-m/ECC	275k	—	~405	Automated safety check: Pass	MIT	3 days ago
Canary Watchaffaan-m/ECC	275k	—	~414	Automated safety check: Pass	MIT	3 days ago

Similar skills

Python Background Jobs
wshobson/agents
Python background job patterns including task queues, workers, and event-driven architecture.
40k GitHub stars~1.8k tokensUpdated 3 days ago
Backend & APIsAuto-check passed
Background
jeremylongshore/tons-of-skills-marketplace
A skill your agent uses when the user wants to see, inspect, cancel, or prune background agents fired during prior chain runs.
2.8k GitHub stars~2.2k tokensUpdated today
Auto-check passed
Canary Watch
affaan-m/ECC
A skill your agent uses to monitor and verify a deployed URL after releases — checks HTTP endpoints, SSE streams, static assets, console errors, and performance regressions after deploys, merges, or…
275k GitHub starsUsed in 1 repo~770 tokens
DevOps & CloudAuto-check passed
Canary Watch
affaan-m/ECC
このスキルを使用して、デプロイメント、マージ、または依存関係アップグレード後にデプロイされたURLの回帰を監視します. An agent skill from affaan-m/ECC.
275k GitHub stars~405 tokensUpdated 3 days ago
DevOps & CloudAuto-check passed
Canary Watch
affaan-m/ECC
使用此技能在部署、合并或依赖升级后监控已部署的URL是否存在回归问题。
275k GitHub stars~414 tokensUpdated 3 days ago
DevOps & CloudAuto-check passed
Background Watch Hook
avibe-bot/avibe
Use vibe watch to run a managed Harness waiter that returns to the same conversation later.
620 GitHub stars~7.7k tokensUpdated today
Agent WorkflowsAuto-check passed

More from WrongStack/WrongStack

All 38 skills in this repo

Design Craft
WrongStack/WrongStack
Design or substantially improve user-facing interfaces with a product-specific visual direction, content hierarchy, and rendered critique.
370 GitHub stars~2k tokensUpdated yesterday
Auto-check passed
Design Critique
WrongStack/WrongStack
A skill your agent uses to audit an interface that already exists and say precisely why it looks generated, templated, or unfinished — a scored rubric across composition, typography, color, states…
370 GitHub stars~3k tokensUpdated yesterday
Auto-check passed
Mailbox Bridge
WrongStack/WrongStack
A skill your agent uses when external coding agents (Claude Code, Aider, custom scripts) need to participate in the project's shared WrongStack mailbox, or when a user asks to "expose the mailbox"…
370 GitHub stars~3.9k tokensUpdated yesterday
Auto-check passed
Multi Agent
WrongStack/WrongStack
A skill your agent uses whenever work can be split across multiple AI agents running in parallel, or when orchestrating leader/worker patterns in WrongStack.
370 GitHub stars~3.6k tokensUpdated yesterday
Auto-check passed
Web Platform Baseline
WrongStack/WrongStack
Use this skill before asserting that a CSS, HTML or accessibility capability is available, unavailable, or the right tool — it carries dated, refreshable platform facts and refuses to let stale…
370 GitHub stars~1.6k tokensUpdated yesterday
Auto-check passed
Wrongstack Mailbox
WrongStack/WrongStack
A skill your agent uses when the user wants to communicate with WrongStack's shared project mailbox from outside WrongStack — read messages sent by WrongStack agents, send replies, broadcast to all…
370 GitHub stars~3.5k tokensUpdated yesterday
Auto-check passed

Questions about Shadow Agent

What does Shadow Agent do?

A skill your agent uses when you need a background monitoring agent that watches the fleet, detects anomalies, and can intervene on command. Shadow Agent is an agent skill from WrongStack/WrongStack. Use this skill when you need a background monitoring agent that watches the fleet, detects anomalies, and can intervene on command.

When should I use Shadow Agent?

Shadow Agent fits situations like: you need a background monitoring agent that watches the fleet; detects anomalies; can intervene on command.

How do I install Shadow Agent in Claude Code?

Run `npx skills add WrongStack/WrongStack --skill shadow-agent -a claude-code`. Or copy the skill folder (.wrongstack/skills/shadow-agent in WrongStack/WrongStack) into .claude/skills/shadow-agent in your project. Claude Code loads it when a task matches its description.

How do I install Shadow Agent in Codex?

Run `npx skills add WrongStack/WrongStack --skill shadow-agent -a codex`. Or copy the skill folder (.wrongstack/skills/shadow-agent in WrongStack/WrongStack) into .agents/skills/shadow-agent in your project. Codex loads it when a task matches its description.

Can I use Shadow Agent in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add WrongStack/WrongStack --skill shadow-agent -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/shadow-agent, .gemini/skills/shadow-agent, .github/skills/shadow-agent and .opencode/skills/shadow-agent in your project.

What does Shadow Agent need to run?

SKILL.md names no scripts, command-line tools or credentials: Shadow Agent is instructions for the agent only.

Does Shadow Agent access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Shadow Agent safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Shadow Agent use?

Shadow Agent is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Shadow Agent use?

About 2.3k tokens (SKILL.md is roughly 9k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Shadow Agent?

Skills that share tags, products or a category with Shadow Agent: Python Background Jobs (wshobson/agents, 40k stars), Background (jeremylongshore/tons-of-skills-marketplace, 2.8k stars), Canary Watch (affaan-m/ECC, 275k stars) and Canary Watch (affaan-m/ECC, 275k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Shadow Agent?

WrongStack (a GitHub organization) maintains it in WrongStack/WrongStack, which has 370 GitHub stars. The repository holds 38 skills in this directory. The repository was last updated on October 7, 2026.

Source: WrongStack/WrongStack on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.

Install Shadow Agent

Facts

At a glance

What it does

When your agent uses it

Example prompts

Workflow steps

What it can do on your machine

Tool permissions

Runs code

Network

Credentials

Context cost

Safety

SKILL.md

Shadow Agent — Fleet Monitoring & Intervention

Overview

Architecture

Core Responsibilities

1. Fleet Monitoring (every 30s by default)

2. Mailbox Surveillance

3. Spike Detection

4. Deterministic Scheduling

5. LLM Analysis

6. Intervention Commands

Data Structures

ShadowState (persisted)

AgentSnapshot

SpikeEvent

Anomaly

Output Format

Status Report

Intervention Log

Workflow

Startup

Heartbeat (every intervalMs)

FleetBus Events

Mailbox Monitoring

Intervention (hoop)

Configuration

Skills in scope

Anti-Patterns

Files

Compare with similar skills

Similar skills

More from WrongStack/WrongStack

Questions about Shadow Agent