soul-guard

soul-guard is a skill for Claude Code, Codex from Parad0x-Labs/openclaw-skills. It costs 46 tokens per session (233 once invoked), scanned B, original, MIT.

A security checker for SOUL.md files, which are text files that define an agent’s identity and instructions.

In plain words
What is it for?
Use it to check an identity file at startup or watch it for changes, especially when other people or systems can edit the file.
Why use it?
It can detect prompt-injection text or unauthorized changes before a modified identity file influences the agent.

Skill for Claude CodeCodex

Which agent this was written for is unclear — built for openclaw. Also seen: built for openclaw.

Good fit Use it to check an identity file at startup or watch it for changes, especially when other people or systems can edit the file.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/parad0x-labs/openclaw-skills/soul-guard
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 Parad0x-Labs/openclaw-skills --skill soul-guard
Clone the repo
git clone --depth 1 https://github.com/Parad0x-Labs/openclaw-skills

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 soul-guard

README.md
[![agentmods](https://agentmods.dev/badge/skills/parad0x-labs/openclaw-skills/soul-guard/github.svg)](https://agentmods.dev/skills/parad0x-labs/openclaw-skills/soul-guard)
Your own site
<a href="https://agentmods.dev/skills/parad0x-labs/openclaw-skills/soul-guard"><img src="https://agentmods.dev/badge/skills/parad0x-labs/openclaw-skills/soul-guard/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 soul-guard

Your own site · 80×15
<a href="https://agentmods.dev/skills/parad0x-labs/openclaw-skills/soul-guard"><img src="https://agentmods.dev/badge/skills/parad0x-labs/openclaw-skills/soul-guard.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 46 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 233 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. 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 $0.00046 $0.00233
Opus 5 $0.00023 $0.00117
Sonnet 5 $0.00009 $0.00047
Haiku 4.5 $0.00005 $0.00023

Measured 12d ago against content hash 07249f80e05b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade B, and why

soul-guard scanned grade B with 1 finding 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 12d ago.

The scan reads SKILL.md. This mod also ships 1 executable file (soul_guard.py), 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.

Instruction-override phrasingmediumPrompt injection

Text telling the model to disregard its earlier instructions or safety rules is the shape of a prompt injection, whoever wrote it.

agents who they are — if one gets tampered with ("ignore previous instructions",

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

plugins/soul-guard/SKILL.md · 28 lines

What it actually says

Soul Guard

Detects prompt injection backdoors in agent identity files. SOUL.md files tell agents who they are — if one gets tampered with ("ignore previous instructions", "you are now..."), Soul Guard catches it before it affects the agent.

Runs as a startup check (once before the agent loads its identity) or a background watcher (re-checks on file change). Alerts on any injection pattern or unauthorized modification.

When to use

  • Any agent that loads its identity from a SOUL.md or similar identity file
  • Shared agent deployments where the identity file could be written by untrusted parties
  • Any agent that needs to assert its own identity hasn't been silently overwritten

Source

github.com/Parad0x-Labs/openclaw-skills/tree/main/plugins/soul-guard

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 12d ago First seen · 28 lines · 46 tokens per session scan B 07249f80e05b

Subscribe to this mod's changes

soul-guard is a skill published in the GitHub repository Parad0x-Labs/openclaw-skills (22 stars, last pushed 9d ago), licensed MIT. It adds 46 tokens to every session and 233 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 1 finding (instruction-override phrasing). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.