allowed-stack-guardrail

A boundary check for conversational product building that compares a requested need with the deployment's approved building blocks.

In plain words
What is it for?
Use it before agreeing to build something for a non-technical person, so supported needs proceed and unsupported needs can be turned into an issue for the relevant project.
Why use it?
It prevents the system from promising work outside what it is set up to operate, while giving unmatched requests a clear next step.

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/devantler-tech/agent-plugins/allowed-stack-guardrail
Any agent
npx skills add devantler-tech/agent-plugins --skill allowed-stack-guardrail
Clone the repo
git clone --depth 1 https://github.com/devantler-tech/agent-plugins

Made for: Claude Code, Codex.

Per session 115 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,124 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 $0.00115 $0.01124
Opus 5 $0.00057 $0.00562
Sonnet 5 $0.00023 $0.00225
Haiku 4.5 $0.00012 $0.00112

Measured yesterday against content hash 6583fb62d212, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

allowed-stack-guardrail 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 yesterday.

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.

plugins/vibe-coding/skills/allowed-stack-guardrail/SKILL.md · 75 lines

How it starts

The opening of the file, as written. The whole thing — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Allowed-stack guardrail

A deployment that lets people build conversationally still has a hard boundary: only the building blocks the deployment actually operates are buildable. This skill is that boundary check. It runs before you agree to build anything, and its verdict is binding: there is no best-effort path around it. Declines are delivered in the register of the jargon-free-voice skill.

The Stack map contract

The allowed stack is deployment-owned configuration — it never ships inside this skill. The consuming deployment's canonical instructions file (AGENTS.md) defines it in a section titled exactly ## Stack map, containing:

  • A table whose rows each carry three required fields:
    • Building block — the block's plain-language name;
    • Good for — the needs it serves, written in the user's vocabulary (this is the matching surface);
    • Owning repoowner/repo, where a suggested issue for that block is filed.
  • A default intake repo (required, once per map) — the catch-all owner/repo that receives the suggested issue for any need matching no row.

Read the section fresh each session; the deployment can change it at any time.

The check

  1. Match conservatively. Compare the user's stated need (outcome / audience / workflow — as elicited by the needs-stack-mapping skill) against each row's Good for purposes. Matching is semantic but conservative: only a confident match counts. Anything you cannot confidently place falls through to the unmatched path — never stretch a row to fit.
  2. In-stack → proceed. A confidently matched need goes forward to be built with the matched block(s), the deployment's way.
  3. Out-of-stack or unmatched → decline and redirect. In plain language: say what you can't build, in one friendly sentence, without technical vocabulary; then offer to put it on the deployment's wish list. With the user's consent, prepare a well-formed issue — the need as the user stated it, the outcome it serves, and why it fell outside the current building blocks. Routing follows the confidence of the match: a need that confidently concerns a mapped block yet is still out-of-stack (say, a capability that block doesn't have) is filed on that block's owning repo; every non-confident or unmatched need goes to the default intake repo — never to a merely "nearest" block, which would land out-of-stack work on an unrelated owner. The user consents to "adding it to the wish list", never to "filing an issue on a repo" — the redirect itself stays jargon-free.

Read the full file on GitHub · 75 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. yesterday First seen · 75 lines · 115 tokens per session scan A 6583fb62d212

Subscribe to this mod's changes

allowed-stack-guardrail is a skill published in the GitHub repository devantler-tech/agent-plugins (2 stars, last pushed yesterday), licensed Apache-2.0. It adds 115 tokens to every session and 1,124 once invoked, about $0.0006 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.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens