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.
npx skills add escoffier-labs/skillet --skill tastegit clone --depth 1 https://github.com/escoffier-labs/skilletWrote 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/escoffier-labs/skillet/taste)<a href="https://agentmods.dev/skills/escoffier-labs/skillet/taste"><img src="https://agentmods.dev/badge/skills/escoffier-labs/skillet/taste/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/escoffier-labs/skillet/taste"><img src="https://agentmods.dev/badge/skills/escoffier-labs/skillet/taste.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 | $0.00048 | $0.01244 |
| Opus 5 | $0.00024 | $0.00622 |
| Sonnet 5 | $0.00010 | $0.00249 |
| Haiku 4.5 | $0.00005 | $0.00124 |
Grade A, and why
taste 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 12d 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.
This is a copy
98% identical to taste — 2 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
taste
You decide what the dish should taste like before you cook it, and nothing leaves the kitchen untasted. The test is the taste: written first, it says what the code should do; written after, it only confirms what the code happens to do. The chef who tastes only at the end learns what the diner was about to find out.
Core principle: if you never watched the test fail, you do not know it tests anything. Violating the letter of this rule is violating its spirit.
The iron law
No production code without a failing test first. Wrote code before the test? Delete it and start over. Not "keep it as reference", not "adapt it while writing the test", not "look at it once". Delete means delete; implement fresh from the test.
Applies to features, bugfixes, refactors, behavior changes. The only exceptions are throwaway prototypes and generated code, and those get the user's explicit sign-off, not your own.
The loop
- RED - write one minimal failing test. One behavior, a name that describes it, real code over mocks. For a bug, the test reproduces the bug; that is how the complaint gets tasted before the dish is re-cooked.
- Watch it fail. Run it, read the output. It must fail, for the expected reason (the feature is missing), not error on a typo. Passes immediately? It tests existing behavior; fix the test.
- GREEN - minimal code to pass. Just enough. No extra options, no adjacent refactoring, no features the test does not demand. YAGNI.
- Watch it pass. Run it, read the output, confirm the rest of the suite stayed green and the output is clean. Fails? Fix the code, never the test.
- REFACTOR on green only. Duplication out, names improved, tests stay green, no new behavior.
- Next behavior, next failing test.
# RED - taste first: this fails with "no attribute dedupe_notes"
def test_dedupe_keeps_first_occurrence(tmp_store):
notes = [{"id": "a", "v": 1}, {"id": "b"}, {"id": "a", "v": 2}]
assert store.dedupe_notes(notes) == [{"id": "a", "v": 1}, {"id": "b"}]
# GREEN - minimal pass, nothing the test didn't ask for
def dedupe_notes(notes):
seen, out = set(), []
for n in notes:
if n["id"] not in seen:
seen.add(n["id"])
out.append(n)
return out
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 12d ago First seen · 83 lines · 48 tokens per session scan A ffc6143e1d85
taste is a skill published in the GitHub repository escoffier-labs/skillet (4 stars, last pushed 11d ago), licensed MIT. It adds 48 tokens to every session and 1,244 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. It is 98% identical to taste, differing in 2 lines, and is treated as a copy.
Other skills, from other repositories
gentle-ai
Use Gentle AI harness discipline for Pi work: clarify first, preserve OpenSpec artifacts, use strict TDD where available, delegate through subagents when useful, and protect review workload.
test-first-bugs
Enforces a test-driven bug-fixing workflow. Use when a user reports a bug, failing code, an error, or asks to fix something.
nl-to-constraints
Transforms natural language requirements (user stories, verbal descriptions, business rules) into formal specifications and constraints. Use when converting informal requirements into structured, testable specifications with explicit constraints. Outputs in multiple formats including BDD-style Given-When-Then, JSON…
lsp-refactoring
Intelligent code refactoring using IDE-level tools (rename, find-references, go-to-definition), AST-aware pattern matching, and TDD verification. Use for safe, large-scale refactoring with precision.
tdd-workflow
A test-first development workflow, commonly called TDD, where a failing test is written before the code that should make it pass. The cycle then adds the smallest working change and cleans it up.
user-story
Create user stories with Mike Cohn format and Gherkin acceptance criteria. Use when turning user needs into development-ready work with clear outcomes and testable conditions.