Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add V-Songbird/slag/plugin install scribeWrote 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/v-songbird/slag/clarify)<a href="https://agentmods.dev/skills/v-songbird/slag/clarify"><img src="https://agentmods.dev/badge/skills/v-songbird/slag/clarify.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 | $0.00118 | $0.00877 |
| Opus 5 | $0.00059 | $0.00439 |
| Sonnet 5 | $0.00024 | $0.00175 |
| Haiku 4.5 | $0.00012 | $0.00088 |
Grade A, and why
clarify scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
clarify — define the task before cutting it
The request could honestly be executed in materially different ways. Your job is to close every one of those gaps before you cut, then execute. Never mention this skill's machinery; just ask well.
If this session is unattended — headless, goal-driven, a background agent — do not ask. Take the most reasonable reading, proceed, and say so in one line when you report. Under a silence-first output style, an ask made here is the permitted case: you are blocked on input only the user can provide.
The rounds
- Check the project's remembered answers. Run
node "${CLAUDE_PLUGIN_ROOT}/scripts/cli.js" memory --jsonfrom the project root — it returns the folded map, so never read.scribe/memory.jsonlyourself. A remembered term is a stated assumption, never a question: fold it into the close line ("Assuming improve = speed here, as remembered") and drop that question from the round. - Blind-spot pass first. Name to yourself the one reading you would act on if nobody were there to ask, then look for what does not fit it: neighboring code, tests, callers, docs, error paths — and at each, whether the thing is wrong, missing, unexplained, or fits so perfectly it deserves a second look. A finding becomes a question only if it changes what you would build.
- Ask the whole frontier in ONE AskUserQuestion call — every question that is currently askable. The tool holds 4, so a wider frontier spills into the next round; lead with the questions whose answers unlock others, then by materiality. Do not drip questions one at a time.
- Every question carries options (max 4), your best guess first, its label suffixed "(Recommended)". Ground each option in this codebase — what the file actually does, who actually calls it — never generic categories.
- Never ask what you can read. Facts derivable from the prompt, the repository, or configuration are not questions. Asking one burns trust.
- Keep asking until the frontier is empty. A follow-up is earned two ways: it spilled over from a call that held only 4, or an answer unlocked it ("Speed" may unlock "the hot path is in a dependency — wrap it or refactor it too?"). Stop when neither is left, not before. Leaving a material gap unasked is the failure this skill exists to prevent.
- Close in one line. State the now-defined task ("Task: memoize
normalize()at this call site, measure before and after") and execute. The user can veto in a glance.
What ships with it
1 file 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.
- 7d ago First seen · 68 lines · 118 tokens per session scan A 65bd28ec15de
clarify is a skill published in the GitHub repository V-Songbird/slag (1 stars, last pushed 3d ago), licensed MIT. It adds 118 tokens to every session and 877 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.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
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…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…
chronicle
Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…
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…