Agent skill · Audited from source

cyrus-setup-launch

skill · ceedaragents/cyrusApache-2.0 licenceAudited 24 Oct 2026 · source 522039659b35

★ 867pushed today
✕BLOCKGrade D · 69/100 · 6 findings · 5220396
867repo stars · whole ceedaragents/cyrus

Is it safe? No — BLOCK, grade D (69/100). We read this skill's source at commit 522039659b35 and found no filesystem, network or shell access and no hard-coded secrets. The audit raised 6 findings, 1 critical or high. What it does: Print a summary of the Cyrus setup and offer to start the agent.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “cyrus-setup-launch (ceedaragents/cyrus)”. Retrieved 24 Oct 2026. https://modelsatlas.com/skills/ceedaragents/cyrus-setup-launch

Blocked. The page stays public so you can check this skill if someone recommended it to you. The audit found 1 critical or high finding, starting with “Privilege escalation / persistence” in SKILL.md. Do not install it without reading them.

2 skills are called “cyrus-setup-launch”. You are viewing ceedaragents/cyrus. 1 of them is BLOCK.

Trust audit

Audit 0.4.1 · source read 24 Oct 2026

Trust score69/100✕BLOCK D
0 critical1 high5 medium0 low0 info

BLOCK means the audit found something that could harm you or your machine. Grade D is shared by 642 of 14,816 audited skills.

What the source does

  • FilesystemNone observedNo reads or writes in the source
  • NetworkNone observedNo outbound calls in the source
  • ShellNone observedNo commands run in the source
  • DependenciesPinnedNothing floats
  • SecretsNone foundNo keys in source

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityWARN5
L1 Static sourceSecrets, path traversal, supply chain, hidden codeNA0
L2 InstructionsPrompt injection, authority framing, hidden directivesFAIL1
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (6)

HIGHPrivilege escalation / persistencereview.unsafe_action · SKILL.md
F-06 Privilege escalation / persistenceCWE-269 · CWE-250Layer L2
sudo security add-trusted-cert -d -r trustRoot -k /Library/Keychains/System.keychain ~/.cyrus/certs/cyrus-egress-ca.pem
Why
Installs a root CA into the system keychain with sudo, bypassing normal user trust, to make an egress proxy CA globally trusted.
Fix
rewrite it so the instruction says plainly what it does, and asks the user before it acts

Audit 0.4.1 · source sha 522039659b35 · 24 Oct 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publisherceedaragents GitHub organizationOwner of the repository this skill is read from
  • Source folderceedaragents/cyrus/skills/cyrus-setup-launchThe exact folder we audited
  • LicenceApache-2.0As the repository states it
  • Commit pinned for this verdict522039659b35A new audit runs when the source changes
  • Repository statusNot archivedPushed today
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: Print a summary of the Cyrus setup and offer to start the agent.

---
name: cyrus-setup-launch
description: Print a summary of the Cyrus setup and offer to start the agent.
---

**CRITICAL: Never use `Read`, `Edit`, or `Write` tools on `~/.cyrus/.env` or any file inside `~/.cyrus/`. Use only `Bash` commands (`grep`, `printf >>`, etc.) to interact with env files — secrets must never be read into the conversation context.**

# Setup Launch

Prints a summary of the completed setup and offers to start Cyrus.

## Step 1: Gather Configuration

Read current state:

```bash
# Base URL
grep '^CYRUS_BASE_URL=' ~/.cyrus/.env 2>/dev/null | cut -d= -f2-

# Linear
grep -c '^LINEAR_CLIENT_ID=' ~/.cyrus/.env 2>/dev/null

# GitHub
gh auth status 2>&1 | head -1

# Slack
grep -c '^SLACK_BOT_TOKEN=' ~/.cyrus/.env 2>/dev/null

# Repositories
cat ~/.cyrus/config.json 2>/dev/null

# Claude auth
grep -c -E '^(ANTHROPIC_API_KEY|CLAUDE_CODE_OAUTH_TOKEN)=' ~/.cyrus/.env 2>/dev/null
```

## Step 2: Print Summary

Print a formatted summary:

```
┌─────────────────────────────────────┐
│         Cyrus Setup Complete        │
├─────────────────────────────────────┤
│                                     │
│  Endpoint: https://your-url.com     │
│  Claude:   ✓ API key configured     │
│                                     │
│  Surfaces:                          │
│    Linear:  ✓ Workspace connected   │
│    GitHub:  ✓ CLI authenticated     │
│    Slack:   ✓ Bot configured        │
│                                     │
│  Repositories:                      │
│    • yourorg/yourrepo               │
│    • yourorg/another-repo           │
│                                     │
└─────────────────────────────────────┘
```

Use ✓ for configured items and ✗ for skipped/unconfigured items.

## Step 3: Make Cyrus Persistent

Cyrus needs to run as a background process so it stays alive and restarts after reboots. **Use the `AskUserQuestion` tool if available** to ask:

> **How would you like to keep Cyrus running in the background?**
>
> 1. **pm2** (recommended) — Node.js process manager. Simple to set up, auto-restarts on crash, log management built in. Best for most users.
> 2. **systemd** (Linux only) — OS-level service manager. Starts on boot automatically, managed with `systemctl`. Best for dedicated Linux servers.
> 3. **Neither** — just run `cyrus` in the foreground for now (you can set up persistence later).

### Option 1: pm2

The agent should run all of these commands directly:

1. Check if pm2 is installed (`which pm2`). If not, install it (`npm install -g pm2`).
2. Start Cyrus: `pm2 start cyrus --name cyrus`
3. Save the process list: `pm2 save`
4. Run `pm2 startup` — this prints a system-specific command. The agent should run that output command too (it typically requires `sudo`).

After setup, inform the user of useful commands:
- `pm2 logs cyrus` — view logs
- `pm2 restart cyrus` — restart
- `pm2 stop cyrus` — stop

### Option 2: systemd (Linux only)

The agent should run all of these commands directly:

1. Resolve the actual values for the service file:
   ```bash
   CYRUS_BIN=$(which cyrus)
   CYRUS_USER=$(whoami)
   ```

2. Write the service file:
   ```bash
   sudo tee /etc/systemd/system/cyrus.service > /dev/null << EOF
   [Unit]
   Description=Cyrus AI Agent
   After=network.target

   [Service]
   Type=simple
   User=$CYRUS_USER
   EnvironmentFile=/home/$CYRUS_USER/.cyrus/.env
   ExecStart=$CYRUS_BIN
   Restart=always
   RestartSec=10

   [Install]
   WantedBy=multi-user.target
   EOF
   ```

3. Enable and start:
   ```bash
   sudo systemctl daemon-reload
   sudo systemctl enable cyrus
   sudo systemctl start cyrus
   ```

After setup, inform the user of useful commands:
- `sudo systemctl status cyrus` — check status
- `sudo journalctl -u cyrus -f` — view logs
- `sudo systemctl restart cyrus` — restart

### Option 3: Foreground

Run directly:

```bash
cyrus
```

## Step 4: Start ngrok (if applicable)

If the user configured ngrok in the endpoint step, the agent should start it:

```bash
ngrok start cyrus
```

If using pm2, also make ngrok persistent:

```bash
pm2 start "ngrok start cyrus" --name ngrok
pm2 save
```

## Step 5: Sandbox CA Certificate Trust (if sandbox enabled)

If the user's `~/.cyrus/config.json` has `sandbox.enabled: true`, check whether the egress proxy CA certificate is trusted in the system keychain.

**Check if sandbox is enabled:**

```bash
grep -o '"enabled":\s*true' ~/.cyrus/config.json 2>/dev/null | head -1
```

If sandbox is enabled, check trust status:

```bash
# macOS — check System keychain for the Cyrus CA
security find-certificate -c "Cyrus Egress Proxy CA" /Library/Keychains/System.keychain 2>&1
```

- If the cert is found (exit code 0): report ✓ trusted. Offer to set `sandbox.systemWideCert: true` in config.json to skip per-session cert env vars.
- If not found (exit code 44): inform the user and offer to run the trust command:

```bash
sudo security add-trusted-cert -d -r trustRoot -k /Library/Keychains/System.keychain ~/.cyrus/certs/cyrus-egress-ca.pem
```

On Linux, check with `test -f /usr/local/share/ca-certificates/cyrus-egress-ca.crt`. If not present:

```bash
sudo cp ~/.cyrus/certs/cyrus-egress-ca.pem /usr/local/share/ca-certificates/cyrus-egress-ca.crt
sudo update-ca-certificates
```

After trusting system-wide, offer to set `sandbox.systemWideCert: true` in config.json. This skips per-session cert env vars (`NODE_EXTRA_CA_CERTS`, `GIT_SSL_CAINFO`, etc.) since the OS cert store handles trust for all tools.

If the user declines system-wide trust, Cyrus still works — it sets cert env vars per-session. But some tools (Bun, .NET, curl on macOS with SecureTransport) will only work with system-wide trust.

## Step 6: Verify Running

Once Cyrus starts, verify it's listening:

```bash
curl -s http://localhost:3456/status
```

Should return `{"status":"idle"}` or similar.

> Then try assigning a Linear issue to Cyrus, or @mentioning it in Slack, to verify the full pipeline works!

## Completion

> ✓ Cyrus is running and ready. Assign a L

Excerpt, first 6,000 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit 5220396

git clone https://github.com/ceedaragents/cyrus
git -C cyrus checkout 522039659b35

The author documents no install command, so this uses the repository directly. Our verdict covers commit 522039659b35 only.

Works in

The author has not said which agents this skill supports, and we have not tested it in any.

Audit history

Every audit, every source change

  1. ✕BLOCK D 69 / 100

    first audit · source 522039659b35

Similar skills

Questions about cyrus-setup-launch

What does the cyrus-setup-launch skill do?
The Claude Code background agent for Linear, Slack, Github, GitLab etc. you deploy anywhere. Supports Codex, Cursor, Gemini, and Opencode harnesses too.
Is cyrus-setup-launch safe to install?
No — not without reading the findings first. The audit graded it D (69/100) and found 1 critical or high issue in the source. Each one is listed on this page with the file and line it is on.
What can cyrus-setup-launch access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
What do I need installed to use cyrus-setup-launch?
Its own instructions reference sudo. Dependencies are pinned to exact versions.
How current is this page?
The grade is for one exact copy of the source (522039659b35), read on 2026-10-24. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.