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/saadshahd/moo.md/reducenpx skills add saadshahd/moo.md --skill reducegit clone --depth 1 https://github.com/saadshahd/moo.mdWrote 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/saadshahd/moo.md/reduce)<a href="https://agentmods.dev/skills/saadshahd/moo.md/reduce"><img src="https://agentmods.dev/badge/skills/saadshahd/moo.md/reduce.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 5.1 | $0.00032 | $0.00661 |
| Opus 5 | $0.00016 | $0.00331 |
| Sonnet 5 | $0.00006 | $0.00132 |
| Haiku 4.5 | $0.00003 | $0.00066 |
Grade A, and why
reduce 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 6d 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 — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Extract
From the conversation/context/user:
- The text as it stands — the rule, doc, prompt, or answer to be reduced.
- What it must cover — the cases, points, or duties it exists for: stated in the context, or recovered by you from it.
Gate
Proceed only when: the text is already written, what it must cover is in hand, and the set is closed — nothing unseen is coming.
Anything else, say which in one plain line — never reduce anyway:
- Nothing is written yet → there is nothing to reduce.
- What it must cover can't be stated or recovered → every cut is blind — no way to say what a shorter version must keep doing.
- Something unseen will meet it — a real case shows a class no rule yet decides, or the rule decides only the cases it came from → use widen skill.
Derive
Do not trim the text. Deleting sentences keeps whichever ones were already there — not the shortest way to cover the set. Set the old wording aside, take what it must cover, and derive the shortest statement that covers all of it.
If what it must cover has moved since the text was written — cases now in hand it never faced, an objective it never served — the old wording is void twice over: derive from the coverage as it stands today, and what comes out is a new derivation of the current set, never the old statement refit to it.
Between candidate statements, prefer the shortest one that covers everything over the one that keeps the original's shape.
The whole run — derivation, shortening, recovery — happens inside this turn; the first thing handed back is the finished result.
Shorten
Push the derived statement shorter, pass by pass. When a pass makes it stop covering something, it went too far — back off by the minimum that recovers it. The overshoot is how you find the floor; what ships covers everything.
The passes end at the floor: every remaining cut costs coverage, and its minimal back-off restores the words that were there. A clause that survives every pass is load-bearing, not fat.
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.
- 6d ago First seen · 51 lines · 32 tokens per session scan A 370cd99bdb90
reduce is a skill published in the GitHub repository saadshahd/moo.md (34 stars, last pushed 20d ago), licensed MIT. It adds 32 tokens to every session and 661 once invoked, about $0.0002 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 2026-08-30.
Other skills, from other repositories
session-recap
Document Claude Code sessions by extracting knowledge into cross-referenced documentation. Triggers on "recap the session", "summarize the work", or after significant code changes.
knowledge-bank-lookup
Delegates knowledge bank lookups to Explore subagents with reflections-first strategy that learns from past mistakes. Checks /reflections/ directory before documentation to extract documented failures to avoid and proven approaches to apply. Automatically triggers when user mentions services ([project-a], [project-b]…
session-manager
Manage current Claude Code session metadata — set project, tags, and summary on session.md via Obsidian CLI. Use this skill whenever the user wants to tag, categorize, name, label, summarize, or organize the current session. Triggers on phrases like "tag this session", "this was about X", "mark this as", "session…
lark-workflow-standup-report
日程待办摘要:编排 calendar +agenda 和 task +get-my-tasks,生成指定日期的日程与未完成任务摘要。适用于了解今天/明天/本周的安排。.
kb-lint
Health-check the knowledge bank for orphan documents, stale content, broken WikiLinks, index drift, and missing frontmatter. Generates a severity-graded report. Use this skill whenever the user mentions lint, health check, KB maintenance, stale docs, broken links, orphan pages, missing references, knowledge bank…
conversation-export
This skill should be used when the user asks to "save the conversation", "export chat", "save chat history", "save our discussion", "export this conversation", "save dialogue", "save this session", "export as markdown", "export as HTML", "summarize and save", "append to chat", "continue saving", "search…