Reins Runtime Security

What it does

When your agent uses it

• Adding guardrails so an agent cannot run irreversible or destructive commands
• Keeping an audit trail of every shell, file and MCP action an agent takes
• Protecting credential folders and sensitive paths from agent writes
• Choosing a permissive, balanced or strict security level for agent sessions

Example prompts

• “Set up Reins in this project with the strict security level.”
• “Why was my git push --force blocked, and how can I allow it for this repo?”
• “Run a reins scan and tell me what it finds in this environment.”
• “Show me the recent decisions the hooks have logged.”

Requirements

• npm, to install the reins CLI globally
• Hook support through .claude/settings.json
• A Watchtower connection, only for centralized policy and audit

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

1. 1Watchtower org policies — CRITICAL rules are immutable
2. 2Watchtower team policies
3. 3Local overrides at ~/.openclaw/reins/policy.json
4. 4Built-in defaults (balanced: reads ALLOW, writes ASK, deletes DENY)

What it can do on your machine

Read from SKILL.md and the folder at commit d5bc85d. 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

  Ships 1 file in scripts/ (Shell), which the agent can run.

  Shell commands in SKILL.md call:

  • git
  • npm
  • bash
  • sh
  • pip

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md. Its commands use git, npm and pip, which can reach the network depending on how they are called.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names these keys or tokens, usually read from environment variables:

  • REINS_WATCHTOWER_API_KEY

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Reins Runtime Security loads about 1.4k tokens when it runs. Until then it costs about 115 tokens; SKILL.md has 553 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~115

When it runs · the whole SKILL.md, loaded when a task matches

~1.4k

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

• WarningMentions a credentials file (SSH keys, cloud or package-manager tokens)SKILL.md:54

  - Writes blocked to protected paths: `~/.ssh`, `~/.gnupg`, `~/.env`,

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); the scripts in this folder are not scanned.

SKILL.md

The full file from pegasi-ai/reins at commit d5bc85d, republished under its Apache-2.0 licence (© pegasi-ai). 553 words, ~1,448 tokens.

name

reins

description

Use this skill whenever security, policies, governance, guardrails, compliance, or safety are relevant — including blocked commands, audit trails, dangerous operations, deletions, file modifications, shell commands, MCP access, API calls, network requests, credentials, or any action that could be irreversible or destructive. Reins installs deterministic PreToolUse/PostToolUse hooks and connects to Watchtower (app.pegasi.ai) for centralized governance.

reins --version

npm install -g @pegasi/reins && reins init

reins init                 # Setup wizard: hooks + policy + Watchtower
reins status               # Show hook and Watchtower connection status
reins sync                 # Pull latest policies from Watchtower; flush pending audit entries
reins policy               # View and edit security policy interactively
reins stats                # Enforcement counts (allowed / blocked / approved)
reins audit -n 20          # Last 20 audit decisions
reins scan                 # Security scan for misconfigurations
reins scan --monitor       # Diff against saved baseline, alert on drift
reins disable              # Temporarily suspend all enforcement
reins enable               # Resume enforcement
reins upgrade              # Pull latest version from npm

REINS_WATCHTOWER_API_KEY=wt_...
REINS_WATCHTOWER_BASE_URL=https://app.pegasi.ai  # default

{"timestamp":"2026-04-15T22:39:42Z","module":"Shell","method":"bash","decision":"BLOCKED","reason":"critical: rm -rf /","tool":"Bash","decisionTime":12}

© pegasi-ai, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file (scripts) in skill of pegasi-ai/reins.

• SKILL.md
• scripts/ensure_installed.sh

Open the folder on GitHubat commit d5bc85d