A guided way to turn a newly discovered coding mistake, or “footgun,” into a proposed rule for the Quellis tool. It shows the proposed trigger for review but does not activate it.

Use it to describe a mistake, run the Quellis teaching command, and review the resulting trigger before deciding whether to enable it.

It avoids having to write a trigger file by hand during a coding session, while keeping activation under your control.

teach — Skill for Claude Code by ondrej-svec/heart-of-gold-toolkit

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 ondrej-svec/heart-of-gold-toolkit --skill teach

Clone the repo

git clone --depth 1 https://github.com/ondrej-svec/heart-of-gold-toolkit

Made for: Claude Code.

Or install quellis-architect, the plugin that ships this one along with the rest of its 3 skills.

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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/ondrej-svec/heart-of-gold-toolkit/teach/github.svg)](https://agentmods.dev/skills/ondrej-svec/heart-of-gold-toolkit/teach)

Your own site

<a href="https://agentmods.dev/skills/ondrej-svec/heart-of-gold-toolkit/teach"><img src="https://agentmods.dev/badge/skills/ondrej-svec/heart-of-gold-toolkit/teach/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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/ondrej-svec/heart-of-gold-toolkit/teach"><img src="https://agentmods.dev/badge/skills/ondrej-svec/heart-of-gold-toolkit/teach.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 69 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,586 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits

NVIDIA SkillSpector pass 7 Sept 2026

How audits are shown

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.

Model	Per session	Once invoked
Fable 5.1	$0.00069	$0.01586
Opus 5.5	$0.00028	$0.00634
Sonnet 5.5	$0.00014	$0.00317
Haiku 4.5	$0.00007	$0.00159

Measured 2mo ago against content hash c74b67d43279, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

teach 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 2mo 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.

plugins/quellis-architect/skills/teach/SKILL.md · 138 lines

How it starts

The opening of the file, as written. The whole thing — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Teach (capture a footgun as a proposed trigger)

A guided wrapper around quellis teach. The CLI exists; this skill exists because most users won't drop to a terminal mid-session to write a TOML trigger by hand. Walks the user through naming the finding, runs the CLI, shows the proposed pack entry, and stops — activation is a deliberate second step.

Boundaries

This skill MAY: ask the user to describe the finding, run quellis teach --finding "<text>", read the proposed trigger output, write a one-line summary to the conversation. This skill MAY NOT: auto-activate the proposed trigger (the CLI's --dry-run-and-review contract is the whole point), edit existing triggers, run other quellis subcommands without explicit user approval, write to the live pack outside of what quellis teach writes.

Teach never auto-promotes. The §2.D acceptance-log substrate exists so triggers earn promotion through observed use; bypassing that throws away the compounding loop.

Common Rationalizations

Shortcut	Why it fails	The cost
"Skip the dry-run; just write the trigger directly"	The CLI's classification step produces canonical DON'T/TRIGGER/VERIFICATION/ORIGIN form. Hand-rolled triggers drift from that form and don't compound cleanly into V1.2 personalization.	Pack inconsistency; harder to review patterns later.
"Activate immediately because the user said yes"	"Activate immediately" is what §2.D told us doesn't work — the calibration step matters. Proposed triggers sit until the user reviews the pack.	Trigger fatigue from over-eager additions.
"Capture every annoying thing as a trigger"	Triggers are for footguns, not preferences. A finding worth a trigger involves real risk — security, destructive, integrity.	Pack noise; real triggers lose signal.
"Skip the why — just record the don't"	The CLI writes `ORIGIN: <why>` for a reason. The why is what calibrates later — was this an incident, a near-miss, a hunch?	Future-you can't tell if a trigger still earns its keep.

Read the full file on GitHub · 138 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.

2mo ago First seen · 138 lines · 69 tokens per session scan A c74b67d43279

Subscribe to this mod's changes

teach is a skill published in the GitHub repository ondrej-svec/heart-of-gold-toolkit (19 stars, last pushed 1mo ago), licensed MIT. It adds 69 tokens to every session and 1,586 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-08-30.

Learning and teaching

teach next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
teachondrej-svec/heart-of-gold-toolkit	Skill	19	1mo ago	MIT	A	69	Claude Code
hr-onboardingnexu-io/open-design	Skill	106k	today	Apache-2.0	A	62	Claude Code, Codex
book-mirrorgarrytan/gbrain	Skill	32k	today	MIT	A	138	Claude Code
exam-forecastanthropics/claude-for-legal	Skill	9.9k	today	Apache-2.0	A	73	Claude Code
lean-formalizewanshuiyin/Auto-claude-code-research-in-sleep	Skill	19k	yesterday	MIT	A	81	Claude Code
eli5companion-inc/feynman	Skill	10k	6d ago	MIT	A	63	Claude Code, Codex
deck-course-modulenexu-io/html-anything	Skill	9.4k	today	Apache-2.0	A	25	Claude Code, Codex