Borrowing it
Nothing to install: this file belongs to hs737/mcp-server-for-ynab. 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/hs737/mcp-server-for-ynab/master/.agents/skills/agent-runtime-guardrails/SKILL.mdgit clone --depth 1 https://github.com/hs737/mcp-server-for-ynabWrote 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/hs737/mcp-server-for-ynab/agent-runtime-guardrails)<a href="https://agentmods.dev/skills/hs737/mcp-server-for-ynab/agent-runtime-guardrails"><img src="https://agentmods.dev/badge/skills/hs737/mcp-server-for-ynab/agent-runtime-guardrails/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/hs737/mcp-server-for-ynab/agent-runtime-guardrails"><img src="https://agentmods.dev/badge/skills/hs737/mcp-server-for-ynab/agent-runtime-guardrails.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00021 | $0.00554 |
| Opus 5 | $0.00010 | $0.00277 |
| Sonnet 5 | $0.00004 | $0.00111 |
| Haiku 4.5 | $0.00002 | $0.00055 |
Grade A, and why
agent-runtime-guardrails 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 9d 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.
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.
How it starts
The opening of the file, as written. The whole thing — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Runtime Guardrails
Use this skill when working on MCP servers, tool execution, planning loops, memory, checkpoints, approvals, or other agent-facing runtime behavior in this Python repository.
Use When
- Adding or editing MCP tools, resources, or prompts
- Changing tool-calling or error-mapping behavior
- Adding memory or checkpoint systems
- Adding human-in-the-loop approval boundaries
- Reviewing runtime safety and observability
Read First
AGENTS.md(if present)README.md- MCP SDK usage in this repo (server setup, lifespan, tool registration)
- Any runtime or workflow docs (for example
docs/agent-design.md,docs/current-state.md)
Core Rules
- Prompts and tool descriptions do not replace durable business rules or validation in code.
- Ephemeral conversation or session state does not replace persistent storage when durability is required.
- Runtime orchestration should be observable (structured logging, clear error types, traceable tool results).
- Human approvals or intervention boundaries should be explicit where relevant.
- Tools must not silently bypass application invariants (budget scope, auth, rate limits).
- Long-running or multi-step behavior should use explicit checkpoints or idempotent steps—not implicit “continue from chat” assumptions.
- Tool and server capabilities advertised to clients must match what is implemented.
- MCP boundary: Each tool should have a narrow, documented contract (Pydantic models or typed parameters). Prefer structured errors (
isError, clear messages) over opaque stack traces in tool results.
Workflow
- Identify what belongs in MCP wiring versus service/domain logic.
- Make durable state transitions explicit in code, not only in prompts.
- Ensure important steps are observable through logs or persisted records.
- Check auth, env vars, and secrets handling (
YNAB_*, tokens via env—not hardcoded). - Update docs if tool behavior or required env changed.
- Verify unhappy paths: API failures, partial data, timeouts, invalid user input.
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.
- 9d ago First seen · 57 lines · 21 tokens per session scan A 660f6918155a
agent-runtime-guardrails is a skill published in the GitHub repository hs737/mcp-server-for-ynab (1 stars, last pushed yesterday), licensed Apache-2.0. It adds 21 tokens to every session and 554 once invoked, about $0.0001 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 2026-08-31.
Other skills, from other repositories
webmcp-nexus
A coding guide for adding WebMCP tools to JavaScript or TypeScript projects. WebMCP lets AI agents call selected application functions through a standard interface.
dark-memory
Use for governance, memory, drift detection, and audit trail via dark-memory-mcp. Covers 52 canonical + 3 red-team tools across 16 namespaces: session lifecycle, agentmemory CRUD+search, vibe-flow spec/artifact publish + drift, LLM-as-judge, delegation+mindset, research, observability, error observatory, governance…
driving-unreal
Use when driving an Unreal Engine 5.x editor through the unreal-ai-connection MCP tools — building or populating a level, spawning/transforming actors, authoring materials and material instances, setting up lighting and atmosphere, reconstructing a scene from a reference photo, setting up a sequencer/cinematic shot…
youtube-transcripts
Read and search YouTube videos through VidWords. Use when the user supplies a YouTube URL, a bare 11-character video id, a youtu.be or Shorts link, or a channel handle, and wants any of: the transcript or subtitles (TXT/SRT/VTT), what a video says about a topic, a quote with a citable timestamp, only the words spoken…
lyrenth-web-reading
Read the content of one or many public web pages through Lyrenth, as clean Markdown instead of raw HTML. Use when the user gives a URL to read, summarize, quote, compare or extract from, or when a link needs to be opened to answer a question.
ml-experiment-review
Use this skill when reviewing an ML experiment, benchmark, ablation, or training run.