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 agentmods add skills/utk2103/prompt-studio/lean-helpnpx skills add utk2103/Prompt-Studio --skill lean-helpgit clone --depth 1 https://github.com/utk2103/Prompt-StudioWrote 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/utk2103/prompt-studio/lean-help)<a href="https://agentmods.dev/skills/utk2103/prompt-studio/lean-help"><img src="https://agentmods.dev/badge/skills/utk2103/prompt-studio/lean-help.svg" alt="Measured on agentmods" 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 | $0.00074 | $0.01439 |
| Opus 5 | $0.00037 | $0.00720 |
| Sonnet 5 | $0.00015 | $0.00288 |
| Haiku 4.5 | $0.00007 | $0.00144 |
Grade B, and why
lean-help scanned grade B with 1 finding 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 4d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
| Check current | `cat ~/.claude/.lean-active` | How it starts
The opening of the file, as written. The whole thing — 142 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Lean Help
Display this reference card when invoked. One-shot only — do NOT change
mode, write flag files, touch ~/.claude/.lean-active, or persist state of
any kind. If the user wants to switch modes, point them at the trigger.
Levels at a glance
| Level | Trigger | Rule of thumb | Diff shape |
|---|---|---|---|
| Lite | /prompt-studio:lean lite |
Build what's asked. Name the lazier alternative in one line. | Requested code + 1-line footnote. |
| Full | /prompt-studio:lean (default) |
The ladder: YAGNI → reuse → stdlib → native → installed dep → one line → minimum. | Shortest working diff. ≤3 lines of prose. |
| Ultra | /prompt-studio:lean ultra |
YAGNI extremist. Deletion beats addition. Challenges the requirement before building. | Often zero code + a question back. |
Level sticks until you change it or the session ends. Persisted in
~/.claude/.lean-active so it survives across turns.
Same task, three levels
Task: "Add a cache for these API responses."
| Level | Response |
|---|---|
| Lite | Custom TTLCache class as asked. Note: functools.lru_cache(maxsize=1000) covers it in one line. |
| Full | @lru_cache(maxsize=1000) on the fetch function. Skipped custom cache, add when lru_cache measurably falls short. |
| Ultra | Do these responses actually re-fetch? Show me a hot path. If yes: @lru_cache. If not: skip. |
Skills
| Skill | Trigger | What it does |
|---|---|---|
| lean | /prompt-studio:lean [lite|full|ultra] |
Lazy mode itself. Shortest solution that actually works. |
| compress | /prompt-studio:compress FILE |
Compress a memory file (CLAUDE.md, todos, prefs) into Lean shorthand. Backup at FILE.original.md. |
| lean-audit | /prompt-studio:lean-audit |
Whole-repo over-engineering audit. Ranked list of what to delete/simplify across the codebase. |
| lean-commit | /prompt-studio:lean-commit |
Generate a terse Conventional Commits message for staged changes. Message only — never runs git. |
| lean-debt | /prompt-studio:lean-debt |
Harvest lean: comments into a debt ledger so deferrals get tracked, not forgotten. |
| lean-gain | /prompt-studio:lean-gain |
Measured impact scoreboard — less code, more speed (no cost claim). |
| lean-review | /prompt-studio:lean-review |
Review a diff for over-engineering only. One line per finding: what to cut, what replaces it. |
| lean-help | /prompt-studio:lean-help |
This card. |
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.
- 4d ago First seen · 142 lines · 74 tokens per session scan B 94ab958c79f9
lean-help is a skill published in the GitHub repository utk2103/Prompt-Studio (7 stars, last pushed 4d ago), licensed Apache-2.0. It adds 74 tokens to every session and 1,439 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
avoid-ai-writing
Audit and rewrite content to remove AI writing patterns ("AI-isms"). Use this skill when asked to "remove AI-isms," "clean up AI writing," "edit writing for AI patterns," "audit writing for AI tells," or "make this sound less like AI." Supports a detect-only mode, an edit-in-place mode for files, an optional voice…
running-long-background-jobs
Use before launching any long-running, backgrounded, detached, or nohup'd process (backfills, batch migrations, bulk data processing, multi-hour scrapes, crawls), and again when verifying it finished correctly. Also use when reconciling a batch job against a corpus (files, DB rows, records) that other processes …
distill-the-scar
Use when you just hit a real failure worth keeping — a bug, a wrong fix, a wasted hour — and want to turn it into a reusable skill instead of a private lesson. Turns one concrete failure into a Whetstone-shaped SKILL.md: extract the transferable invariant, write it so an agent can apply it without the original…
test-fixture-realism
Use when writing tests against a shared store or with fabricated fixtures — a test only proves something if its fixtures mirror data reality and its assertions can't pass trivially. Covers asserting on rows you created (not aggregate counts), real referential integrity, seeded randomness and pinned time, fail-closed…
a-memory-is-not-a-trigger
Use when you're relying on a stored memory, a recalled note, an injected session-start banner, or a description-matched skill to make a required action happen every session or run — especially a recurring logging, bookkeeping, or guardrail step. Passive context is advisory: the agent reads it and can still skip the…
diff-each-side-against-merge-base
Use when resolving a merge or rebase conflict, or reviewing a merge commit — conflict markers show where git gave up, not the full semantic delta, and 3-way auto-merge silently drops fields outside the markers. Compare ours/theirs/base by symbol and field inventory before trusting the result. Triggers on "merge…