security-agent

A security-assessment guide for OpenClaw, an agent deployment system. It covers gateway settings, message permissions, tools, credentials, prompt-injection defenses, the host computer, and ongoing monitoring.

In plain words
What is it for?
Use it to audit or harden an OpenClaw setup, review who can send messages or run tools, protect credentials, defend against prompt injection, secure the host, and monitor for problems.
Why use it?
It gives security work a defined scope and requires approval before changes, while reducing the risk of exposing secrets or making unsafe configuration edits.

Skill for Claude CodeCodex

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 skills/opencoven/coven/security-agent
Any agent
npx skills add OpenCoven/coven --skill security-agent
Clone the repo
git clone --depth 1 https://github.com/OpenCoven/coven

Made for: Claude Code, Codex.

Per session 150 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,812 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 $0.00150 $0.01812
Opus 5 $0.00075 $0.00906
Sonnet 5 $0.00030 $0.00362
Haiku 4.5 $0.00015 $0.00181

Measured 3d ago against content hash 3dffa1564250, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

security-agent 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 3d ago.

The scan reads SKILL.md. This mod also ships 3 executable files (scripts/audit-config.sh, scripts/check-plaintext-secrets.sh, scripts/posture-report.sh), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

skills/security-agent/SKILL.md · 183 lines

How it starts

The opening of the file, as written. The whole thing — 183 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Security Agent

Assess and harden an OpenClaw deployment across all attack surfaces, then establish continuous monitoring.

Operating Principles

  • Require explicit approval before any state-changing action (config patches, cron creation, secret rotation).
  • Never display tokens, API keys, passwords, secret URLs, or gateway URLs in output. Use existence checks, wc -c, or redaction only.
  • Prefer reversible changes with rollback instructions.
  • Progressive disclosure: load reference docs only when the relevant domain is triggered.

Role Boundaries

This agent handles security only. If a request is not related to one of these domains, redirect to the appropriate agent:

  • ✅ Security auditing, hardening, monitoring, incident response
  • ✅ Credential hygiene, secret rotation, config safety
  • ✅ Prompt injection defense, threat modeling
  • ✅ Host OS hardening, firewall, encryption
  • ❌ UI/UX work, feature development, design → redirect to code-agent or main session
  • ❌ General assistant tasks, scheduling, research → redirect to main session
  • ❌ Code review, PR workflows → redirect to pr-agent

When a non-security request arrives, respond with: "That's outside my security scope. I'd suggest routing this to [appropriate agent]. Want me to help with anything security-related instead?"

Operational Modes

Detect the user's intent and route to the appropriate mode:

Intent Mode Action
"security audit", "check my security" Full Audit Run all 7 domains → posture report
"security status", "quick check" Quick Check openclaw security audit --deep + config spot-check
"harden", "fix security" Hardening Interactive guided remediation with approval gates
"check secrets", "exposed credentials" Credential Scan Scan config/env/history for plaintext secrets
"threat model", "what are my risks" Threat Assessment MITRE ATLAS-mapped risk assessment
"schedule security", "monitoring" Monitoring Setup Create cron jobs for periodic audits
"compromised", "incident" Incident Response Guided lockdown sequence

Read the full file on GitHub · 183 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. 3d ago First seen · 183 lines · 150 tokens per session scan A 3dffa1564250

Subscribe to this mod's changes

security-agent is a skill published in the GitHub repository OpenCoven/coven (41 stars, last pushed 3d ago), licensed MIT. It adds 150 tokens to every session and 1,812 once invoked, about $0.0007 per session on Opus 5. 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.