FC openclaw

OpenClaw development assistant, built by Michel Costa, co-founder of Brabaflow — AI-Native Agency (brabaflow.ai). Use this skill when the user asks about OpenClaw — a self-hosted gateway that connects chat apps (WhatsApp, Telegram, Discord, iMessage, etc.) to AI coding agents. Covers configuration, channels, providers, tools, plugins, deployment, CLI commands, and all aspects of OpenClaw development. 333 pages of verbatim official documentation from docs.openclaw.ai.

Not recommendedcritical or high security findings · low grade F

ClawHub Agent Skills author: Brabaflow v1.0.0 MIT-0 22 files body ≈ 2 386 tokens Open the sourceclawhub.ai↗ analyzed 2 d ago

As a process C 51/100 · Has gaps — weak spots: result and completion, inputs and preconditions, consistency

Integration Discord Telegram WhatsApp Infrastructure AI and agents Writing and documentstype and topics are labelled automatically from the skill text

Runs in: OpenClaw Claude / Claude Code ChatGPT / Codex Gemini CLI Cursor GitHub Copilot / VS Code Qwen Code Kimi Code CLI Z.ai (GLM) DeepSeek (Deep Code) Mistral Vibe LangChain (deepagents) Goose OpenCode

JSON

Technical rating

33/100

safety, quality, tests

Safety 60%: 0
Quality 40%: 82
Run on models: none yet

Process rating

51/100

Has gaps

Result and completion w 14: 0
Inputs and preconditions w 11: 0
Progress reporting w 2: 0

the three weakest of ten parameters · all ten

What is at stake

The skill contains fragments that, in the wrong hands, cost money or data. Below: what the installer risks and what the author should do.

Dangerous commands

If you install

The skill contains commands that delete files, rewrite disks or execute code fetched from the network. The agent may run them without asking if it believes the instructions require it.

For the author

Replace destructive commands with safe equivalents that ask for confirmation, scope them to one folder, and stop piping curl into a shell: pin a version and a checksum.

Exfiltration

If you install

The instructions or scripts send data out: environment variables, keys, file contents, chat history. You may never notice, because the agent performs the upload as "part of the task".

For the author

If the upload is not needed for the task, remove it: catalogs flag such skills and delist them. If it is needed, name the destination explicitly, say what leaves the machine, and give the user a switch.

Instruction override

If you install

The text contains phrases like "ignore previous instructions" or "you are now…". That is an attempt to hijack the agent: it may break your rules, the system limits or company policy.

For the author

An honest skill does not need them: state the role and the rules directly without overriding other instructions. Otherwise catalog scanners and corporate filters will block the listing.

Risky intent medium severity

Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.

If you install

The purpose itself is risky: wallets, browser password stores, offensive security. Even an honest implementation gives the agent access to things that cost money.

For the author

Explain in the description why the access is needed and how it is limited; add tests that show refusals on dangerous requests.

How to improve

Address the high-severity findings: each costs 18 safety points. If one is a false positive, add the rule id to guard.allow in spec.yaml.

For the model run — optional

Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.

Guard findings · 73

high Dangerous commands cmd-persistence docs/02-installation.md:1763

Persistence mechanism (cron / launchd / scheduled task / autorun registry)
```
rm -f ~/Library/LaunchAgents/ai.openclaw.gateway.plist
```
high Exfiltration exfil-read-secret-files docs/03-gateway.md:7068

Reads credential / secret files
```
ssh-copy-id -i ~/.ssh/id_rsa <REMOTE_USER>@<REMOTE_IP>
```
high Dangerous commands cmd-persistence docs/03-gateway.md:7122

Persistence mechanism (cron / launchd / scheduled task / autorun registry)
```
launchctl bootstrap gui/$UID ~/Library/LaunchAgents/ai.openclaw.ssh-tunnel.plist
```
high Dangerous commands cmd-persistence docs/04-channels.md:981

Persistence mechanism (cron / launchd / scheduled task / autorun registry)
```
launchctl unload ~/Library/LaunchAgents/com.user.poke-messages.plist 2>/dev/null || true
```
high Dangerous commands cmd-persistence docs/04-channels.md:982

Persistence mechanism (cron / launchd / scheduled task / autorun registry)
```
launchctl load ~/Library/LaunchAgents/com.user.poke-messages.plist
```
high Exfiltration exfil-webhook-url docs/04-channels.md:3064

Webhook / callback URL commonly used for exfiltration (verify the destination)
```
# Set messaging endpoint to: https://abc1….io/api/messages
```
high Dangerous commands cmd-persistence docs/05-providers.md:2470

Persistence mechanism (cron / launchd / scheduled task / autorun registry)
```
cat > ~/Library/LaunchAgents/com.claude-max-api.plist << 'EOF'
```
high Dangerous commands cmd-persistence docs/05-providers.md:2495

Persistence mechanism (cron / launchd / scheduled task / autorun registry)
```
launchctl bootstrap gui/$(id -u) ~/Library/LaunchAgents/com.claude-max-api.plist
```
high Instruction override en-fake-system-prompt docs/06-tools.md:1658

Fake system prompt injected into content
```
*   `systemPrompt`: full system prompt override.
```
high Dangerous commands cmd-persistence docs/12-platforms-desktop.md:128

Persistence mechanism (cron / launchd / scheduled task / autorun registry)
```
schtasks /create /tn "WSL Boot" /tr "wsl.exe -d Ubuntu --exec /bin/true" /sc onstart /ru SYSTEM
```
high Dangerous commands cmd-pipe-to-shell docs/13-deploy.md:83

Downloads and executes remote code from an unrecognised host (pipe to shell)
```
curl -fsSL https://get.docker.com | sh
```
high Dangerous commands cmd-pipe-to-shell docs/13-deploy.md:473

Downloads and executes remote code from an unrecognised host (pipe to shell)
```
curl -fsSL https://get.docker.com | sudo sh
```

Medium and low: 61

medium Dangerous commands cmd-pipe-to-shell docs/02-installation.md:30

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL --proto '=https' --tlsv1.2 https://openclaw.ai/install.sh | bash
```
vendor-host
medium Dangerous commands cmd-pipe-to-shell docs/02-installation.md:34

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL --proto '=https' --tlsv1.2 https://openclaw.ai/install.sh | bash -s -- --help
```
vendor-host
medium Dangerous commands cmd-pipe-to-shell docs/02-installation.md:38

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL --proto '=https' --tlsv1.2 https://openclaw.ai/install-cli.sh | bash
```
vendor-host
medium Dangerous commands cmd-pipe-to-shell docs/02-installation.md:42

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL --proto '=https' --tlsv1.2 https://openclaw.ai/install-cli.sh | bash -s -- --help
```
vendor-host
medium Dangerous commands cmd-pipe-to-shell docs/02-installation.md:80

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL --proto '=https' --tlsv1.2 https://openclaw.ai/install.sh | bash
```
vendor-host
medium Dangerous commands cmd-pipe-to-shell-known-host docs/02-installation.md:405

Pipe-to-shell installer from a well-known host (still executes remote code)
```
curl -fsSL https://raw.githubusercontent.com/openclaw/openclaw-ansible/main/install.sh | bash
```
medium Dangerous commands cmd-shell-rc docs/02-installation.md:741

Writes to a shell startup file
```
echo 'source ~/.clawdock/clawdock-helpers.sh' >> ~/.zshrc && source ~/.zshrc
```
medium Dangerous commands cmd-privilege docs/02-installation.md:885

Privilege escalation / world-writable permissions
```
sudo chown -R 1000:1000 /path/to/openclaw-config /path/to/openclaw-workspace
```
medium Dangerous commands cmd-pipe-to-shell-known-host docs/02-installation.md:898

Pipe-to-shell installer from a well-known host (still executes remote code)
```
RUN curl -fsSL https://bun.sh/install | bash
```
medium Exfiltration exfil-read-secret-files docs/02-installation.md:906

Reads credential / secret files (Dockerfile instruction (build context, not runtime exfiltration))
```
COPY package.json pnpm-lock.yaml pnpm-workspace.yaml .npmrc ./
```
Dockerfile
medium Dangerous commands cmd-persistence docs/03-gateway.md:7096

Persistence mechanism (cron / launchd / scheduled task / autorun registry) (quoted — discussed, not commanded)
```
Save this as `~/Library/LaunchAgents/ai.openclaw.ssh-tunnel.plist`:
```
quoted
medium Dangerous commands cmd-persistence docs/04-channels.md:942

Persistence mechanism (cron / launchd / scheduled task / autorun registry) (quoted — discussed, not commanded)
```
*   `~/Library/LaunchAgents/com.user.poke-messages.plist`
```
quoted
medium Exfiltration exfil-webhook-url docs/04-channels.md:3063

Webhook / callback URL commonly used for exfiltration (verify the destination) (security demo / example)
```
# Copy the https URL, e.g., https://abc1….io
```
demo
medium Exfiltration net-credential-use docs/04-channels.md:3406

Credential used in a network call (verify the destination is the intended service)
```
curl -H "Authorization: Bearer $TOKEN" \
```
medium Exfiltration net-credential-use docs/04-channels.md:3410

Credential used in a network call (verify the destination is the intended service)
```
curl -H "Authorization: Bearer $TOKEN" \
```
medium Dangerous commands cmd-shell-rc docs/05-providers.md:860

Writes to a shell startup file
```
echo 'export AWS_PROFILE=default' >> ~/.bashrc
```
medium Dangerous commands cmd-shell-rc docs/05-providers.md:861

Writes to a shell startup file
```
echo 'export AWS_REGION=us-east-1' >> ~/.bashrc
```
medium Dangerous commands cmd-persistence docs/10-platforms-macos.md:357

Persistence mechanism (cron / launchd / scheduled task / autorun registry) (quoted — discussed, not commanded)
```
*   `~/Library/LaunchAgents/ai.openclaw.gateway.plist` (or `~/Library/LaunchAgents/ai.openclaw.<profile>.plist`)
```
quoted
medium Dangerous commands cmd-privilege docs/12-platforms-desktop.md:200

Privilege escalation / world-writable permissions
```
sudo tee /etc/wsl.conf >/dev/null <<'EOF'
```
medium Exfiltration exfil-read-secret-files docs/13-deploy.md:224

Reads credential / secret files (Dockerfile instruction (build context, not runtime exfiltration))
```
COPY package.json pnpm-lock.yaml pnpm-workspace.yaml .npmrc ./
```
Dockerfile
medium Dangerous commands cmd-privilege docs/13-deploy.md:473

Privilege escalation / world-writable permissions
```
curl -fsSL https://get.docker.com | sudo sh
```
medium Exfiltration exfil-read-secret-files docs/13-deploy.md:618

Reads credential / secret files (Dockerfile instruction (build context, not runtime exfiltration))
```
COPY package.json pnpm-lock.yaml pnpm-workspace.yaml .npmrc ./
```
Dockerfile
medium Dangerous commands cmd-pipe-to-shell docs/13-deploy.md:1277

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL https://openclaw.ai/install.sh | bash
```
vendor-host
medium Dangerous commands cmd-shell-rc docs/13-deploy.md:1611

Writes to a shell startup file
```
echo 'export PATH="$PATH:$HOME/.local/bin"' >> ~/.zshrc && source ~/.zshrc
```
medium Dangerous commands cmd-pipe-to-shell-known-host docs/13-deploy.md:2058

Pipe-to-shell installer from a well-known host (still executes remote code)
```
curl -fsSL https://tailscale.com/install.sh | sh
```
medium Dangerous commands cmd-pipe-to-shell docs/13-deploy.md:2069

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL https://openclaw.ai/install.sh | bash
```
vendor-host
medium Dangerous commands cmd-privilege docs/13-deploy.md:2358

Privilege escalation / world-writable permissions
```
sudo chmod 600 /swapfile
```
medium Dangerous commands cmd-privilege docs/13-deploy.md:2363

Privilege escalation / world-writable permissions
```
echo '/swapfile none swap sw 0 0' | sudo tee -a /etc/fstab
```
medium Dangerous commands cmd-privilege docs/13-deploy.md:2366

Privilege escalation / world-writable permissions
```
echo 'vm.swappiness=10' | sudo tee -a /etc/sysctl.conf
```
medium Dangerous commands cmd-pipe-to-shell docs/13-deploy.md:2375

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL https://openclaw.ai/install.sh | bash
```
vendor-host
medium Dangerous commands cmd-pipe-to-shell-known-host docs/13-deploy.md:2432

Pipe-to-shell installer from a well-known host (still executes remote code)
```
curl -fsSL https://tailscale.com/install.sh | sh
```
medium Dangerous commands cmd-privilege docs/13-deploy.md:2505

Privilege escalation / world-writable permissions
```
echo 'gpu_mem=16' | sudo tee -a /boot/config.txt
```
medium Dangerous commands cmd-pipe-to-shell-known-host docs/13-deploy.md:2721

Pipe-to-shell installer from a well-known host (still executes remote code)
```
curl -fsSL https://deb.nodesource.com/setup_22.x | bash -
```
medium Dangerous commands cmd-pipe-to-shell-known-host docs/13-deploy.md:2772

Pipe-to-shell installer from a well-known host (still executes remote code)
```
curl -fsSL https://tailscale.com/install.sh | sh
```
medium Risky intent intent-wallet-secrets docs/16-security.md:461

Handles crypto-wallet secrets (seed / mnemonic / private key) — a classic stealer target
```
/(wallet|seed phrase|mnemonic|crypto)/i
```
medium Dangerous commands cmd-pipe-to-shell docs/19-troubleshooting.md:426

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL https://openclaw.ai/install.sh | bash -s -- --install-method git
```
vendor-host
medium Dangerous commands cmd-pipe-to-shell docs/19-troubleshooting.md:450

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL https://openclaw.ai/install.sh | bash
```
vendor-host
medium Dangerous commands cmd-pipe-to-shell docs/19-troubleshooting.md:557

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL --proto '=https' --tlsv1.2 https://openclaw.ai/install.sh | bash -s -- --beta
```
vendor-host
medium Dangerous commands cmd-pipe-to-shell docs/19-troubleshooting.md:561

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL --proto '=https' --tlsv1.2 https://openclaw.ai/install.sh | bash -s -- --install-method git
```
vendor-host
medium Dangerous commands cmd-pipe-to-shell docs/19-troubleshooting.md:590

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host)
```
curl -fsSL https://openclaw.ai/install.sh | bash -s -- --install-method git
```
vendor-host
medium Dangerous commands cmd-shell-rc docs/19-troubleshooting.md:852

Writes to a shell startup file
```
echo 'eval "$(/home/linuxbrew/.linuxbrew/bin/brew shellenv)"' >> ~/.profile
```
medium Dangerous commands cmd-pipe-to-shell-known-host docs/19-troubleshooting.md:1492

Pipe-to-shell installer from a well-known host (still executes remote code)
```
curl -fsSL https://tailscale.com/install.sh | sh
```
low Dangerous commands cmd-privilege docs/02-installation.md:1169

Privilege escalation / world-writable permissions (quoted — discussed, not commanded)
```
*   If `agents.defaults.sandbox.browser.noSandbox` is set, `--no-sandbox` and `--disable-setuid-sandbox` are also appended.
```
quoted
low Dangerous commands cmd-privilege docs/03-gateway.md:1728

Privilege escalation / world-writable permissions (quoted — discussed, not commanded)
```
*   plus `--no-sandbox` and `--disable-setuid-sandbox` when `noSandbox` is enabled.
```
quoted
low Dangerous commands cmd-privilege docs/03-gateway.md:3639

Privilege escalation / world-writable permissions (quoted — discussed, not commanded)
```
*   `--no-sandbox` and `--disable-setuid-sandbox` when `noSandbox` is enabled.
```
quoted
low Dangerous commands cmd-background-process docs/03-gateway.md:7420

Starts a background / autostarted process
```
sudo systemctl enable --now openclaw-gateway[-<profile>].service
```
low Exfiltration exfil-webhook-url docs/04-channels.md:49

Webhook / callback URL commonly used for exfiltration (verify the destination) (placeholder value)
```
curl "https://api.telegram.org/bot<bot_token>/getUpdates"
```
placeholder
low Secrets in code secret-password-literal docs/04-channels.md:4255

Hard-coded password / key literal (may be an example) (placeholder value)
```
accessToken: "oaut…...", // OAuth Access Token (or use OPENCLAW_TWITCH_ACCESS_TOKEN env var)
```
placeholder
low Secrets in code secret-password-literal docs/04-channels.md:4299

Hard-coded password / key literal (may be an example) (placeholder value)
```
accessToken: "oaut…...",
```
placeholder
low Secrets in code secret-password-literal docs/04-channels.md:4351

Hard-coded password / key literal (may be an example) (placeholder value)
```
accessToken: "oaut…...",
```
placeholder
low Secrets in code secret-password-literal docs/04-channels.md:4357

Hard-coded password / key literal (may be an example) (placeholder value)
```
accessToken: "oaut…...",
```
placeholder
low Secrets in code secret-password-literal docs/04-channels.md:4508

Hard-coded password / key literal (may be an example) (placeholder value)
```
accessToken: "oaut…...",
```
placeholder
low Exfiltration exfil-webhook-url docs/08-plugins.md:134

Webhook / callback URL commonly used for exfiltration (verify the destination) (placeholder value)
```
// publicUrl: "https://example.ngrok.app/voice/webhook",
```
placeholder
low Dangerous commands cmd-privilege docs/10-platforms-macos.md:1287

Privilege escalation / world-writable permissions (quoted — discussed, not commanded)
```
*   Remove the override: `sudo rm /Library/Preferences/Logging/Subsystems/ai.openclaw.plist`.
```
quoted
low Dangerous commands cmd-shell-rc docs/13-deploy.md:1938

Writes to a shell startup file (detector / deny-list definition)
```
grep -q 'NODE_COMPILE_CACHE=/var/tmp/openclaw-compile-cache' ~/.bashrc || cat >> ~/.bashrc <<'EOF'
```
detector
low Dangerous commands cmd-shell-rc docs/13-deploy.md:2460

Writes to a shell startup file (detector / deny-list definition)
```
grep -q 'NODE_COMPILE_CACHE=/var/tmp/openclaw-compile-cache' ~/.bashrc || cat >> ~/.bashrc <<'EOF' # pragma: allowlist secret
```
detector
low Dangerous commands cmd-background-process docs/13-deploy.md:2583

Starts a background / autostarted process
```
sudo systemctl enable openclaw
```
low Risky intent intent-offensive-security docs/16-security.md:382

Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
```
#### T-EX…03: Credential Harvesting
```
low Risky intent intent-offensive-security docs/16-security.md:459

Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
```
/(malware|stealer|phish|phishing|keylogger)/i
```
low Dangerous commands cmd-pipe-to-shell docs/18-reference.md:1176

Downloads and executes remote code from an unrecognised host (pipe to shell) (the skill's own vendor host; quoted — discussed, not commanded)
```
*   (Optional) Installer E2E (Docker, runs `curl -fsSL https://openclaw.ai/install.sh | bash`, onboards, then runs real tool calls):
```
vendor-hostquoted

…and 1 more

Files scanned: 22. Evidence is masked. Grey chips explain why severity was lowered.

Against the Agent Skills spec

note frontmatter-key unknown frontmatter key "trigger"
note frontmatter-key unknown frontmatter key "source"

Process rating: all ten parameters 51/100

0Result and completion. Does not say what the result is
0Inputs and preconditions. Does not say what the process needs to start
0Progress reporting. Says nothing while it works
30Running it twice. 11 mutating operations with no state check
40Consistency. Frontmatter name (openclaw) differs from the folder (openclaw-agent-skill)
60Tools and files. Uses tools (bash, web, node) that frontmatter does not declare
65Failures and branches. 3 branches
70When it triggers. States when to use, but not when not to
100Steps. 180 steps
100Execution cost. Instruction body is 2386 tokens

medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing

Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.

Quality signals

+5Description has no quoted example phrases that should trigger the skill
+4Description does not say when NOT to use the skill (false activations)
+3Output format is not stated: the model decides each time
+4No input/output examples
+1No license
+2Single-language instructions
+3Description length 471: enough signal without eating the budget
+4Structure: 24 headings
+3Step-by-step instructions: 180 items
+4Reference files are cited in the instructions (20 of 20)

Quality base 70; lint remarks subtract, signals add up to 100. Result: 82.

External checks

ClawHub: suspicious

This is a documentation-only OpenClaw helper, but its bundled docs include multiple high-impact setup, permission, persistence, and privacy examples that need careful review before use.

LLM: suspicious (medium) · VirusTotal: · 29 May 2026

Tests

— no evals.json

— no spec.yaml

— not run on models

Files · 22

docs/01-core-concepts.md 249K
docs/02-installation.md 73K
docs/03-gateway.md 280K
docs/04-channels.md 256K
docs/05-providers.md 91K
docs/06-tools.md 276K
docs/07-automation.md 81K
docs/08-plugins.md 20K
docs/09-nodes.md 52K
docs/10-platforms-macos.md 59K
docs/11-platforms-mobile.md 9K
docs/12-platforms-desktop.md 6K
docs/13-deploy.md 80K
docs/14-cli.md 144K
docs/15-web-uis.md 22K
docs/16-security.md 33K
docs/17-templates.md 22K
docs/18-reference.md 67K
docs/19-troubleshooting.md 173K
docs/20-experiments.md 88K
skill-card.md 3K
SKILL.md 10K

Frontmatter

trigger: Use when user mentions OpenClaw, openclaw, or asks about: configuring AI agent gateways, connecting chat platforms to AI

author: Brabaflow — AI-Native Agency (brabaflow.ai) | Michel Costa

source: docs.openclaw.ai (333 pages, verbatim extraction, 2026-03-09)

Dates

In catalog since 9 Mar 2026

Updated in catalog 11 May 2026

First seen here 10 Sept 2026

Last analysis 11 Sept 2026, 02:37 UTC

Last seen in crawl 7 h ago

Score history

Date		Score	Safety	Qual.
10 Sept 2026	F	33	0	82

Similar skills

A openclaw ClawHub
A openclaw ClawHub
A openclaw TerminalSkills/skills
A openclaw ClawHub
A openclaw ClawHub
B openclaw ClawHub