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 EthanYoQ/Skill-hub --skill yq-windows-trash-cleanergit clone --depth 1 https://github.com/EthanYoQ/Skill-hubWrote 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/ethanyoq/skill-hub/yq-windows-trash-cleaner)<a href="https://agentmods.dev/skills/ethanyoq/skill-hub/yq-windows-trash-cleaner"><img src="https://agentmods.dev/badge/skills/ethanyoq/skill-hub/yq-windows-trash-cleaner/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/ethanyoq/skill-hub/yq-windows-trash-cleaner"><img src="https://agentmods.dev/badge/skills/ethanyoq/skill-hub/yq-windows-trash-cleaner.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Tool Misuse · line 32 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Anti-Refusal · line 182 Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.Fix: Remove instructions that suppress warnings, disclaimers, or ethical commentary. Let the agent surface safety-relevant caveats to the user.
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 | $0.00046 | $0.03051 |
| Opus 5 | $0.00023 | $0.01525 |
| Sonnet 5 | $0.00009 | $0.00610 |
| Haiku 4.5 | $0.00005 | $0.00305 |
Grade A, and why
yq-windows-trash-cleaner 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 11d 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 — 205 lines — stays where its author put it; the contents beside it link to each section on GitHub.
YQ Windows Trash Cleaner
Run an evidence-first Windows maintenance workflow that is safe to invoke at any time and is not tied to one project, drive, application, or prior incident.
Route the request
Distinguish storage from RAM before collecting data:
- Storage intent: C drive/full disk, rapid GB growth, caches, projects, worktrees, Docker/WSL images, backups, sessions, compression, or cleanup.
- RAM intent: processes consuming memory, paged/nonpaged pool, pool tags, startup storms, long-running workload growth, Fast Startup, or suspected leaks.
- Mixed intent: run the storage and memory tracks separately and reconcile them only in the final report.
Classify authorization:
AUDIT_ONLY: inspect and report. Phrases such as “哪些可以删” do not authorize deletion.PLAN_ONLY: produce exact manifest IDs, commands, recovery, and predicted impact without changing state.APPLY_APPROVED: execute only the candidate IDs the user explicitly approved or an exact safe-clean scope the user explicitly requested after seeing the classification.
Scanning is always read-only. A report, HTML button, color, ignored status, or large size is never authorization.
Safety contract
Protect user work before reclaiming space:
- Never delete documents, downloads, sessions, databases, credentials, models, evidence, backups, dirty/unmerged worktrees, release artifacts, Docker volumes, system files, or unknown directories by category name alone.
- Do not run whole-repository
git clean -fdX, blanketdocker volume prune, or raw recursive worktree deletion. - Do not hand-delete WinSxS, DriverStore, Windows Installer, pagefile, hiberfil, recovery data, or VSS data.
- Do not follow reparse points while measuring or deleting. Resolve every target and reject a target equal to an approved root, outside it, or sharing only a text prefix.
- Keep active projects protected. Read repository instructions and current-state documents; check Git state, worktrees, running processes, locks, manifests, lockfiles, evidence paths, and regeneration commands.
- Prefer semantic maintenance: package-manager prune commands,
git worktree remove, Docker object IDs, official DISM analysis/cleanup, transactional SQLite maintenance, or moving a review item to the Recycle Bin. - Hard deletion is off by default. It requires a fresh explicit decision, exact paths/IDs, full preflight, and an action log.
What ships with it
26 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.
- agents/openai.yaml 320 B
- evals/evals.json 7.9 KB
- evals/files/approved-cleanup-fixture.json 801 B
- evals/files/disk-inventory.json 936 B
- evals/files/memory-longrun-samples.csv 697 B
- evals/files/project-dirty-repo/processes.json 238 B
- evals/files/project-dirty-repo/status.txt 37 B
- evals/files/project-ignored-paths/.gitignore 37 B
- evals/files/whole-computer-project-inventory.json 2.1 KB
- references/cleanup-classification.md 1.5 KB
- references/design-provenance.md 1.1 KB
- references/docker-storage.md 1.4 KB
- references/memory-diagnosis.md 2.2 KB
- references/project-cleanup-classification.md 2.7 KB
- references/quality-gates.md 1.5 KB
- references/storage-classification.md 1.9 KB
- references/whole-computer-project-audit.md 2.4 KB
- scripts/Collect-WindowsHealthSnapshot.ps1 6.9 KB runs code
- scripts/Get-ComputerProjectInventory.ps1 11 KB runs code
- scripts/Get-ComputerStorageInventory.ps1 11 KB runs code
- scripts/Get-DockerStorageAudit.ps1 2.5 KB runs code
- scripts/Get-LongRunMemoryAssessment.ps1 5.1 KB runs code
- scripts/Get-ProjectCleanupAudit.ps1 13 KB runs code
- scripts/Invoke-ApprovedCleanupManifest.ps1 14 KB runs code
- scripts/Measure-PathUsage.ps1 2.1 KB runs code
- scripts/New-StorageAuditReport.ps1 9.0 KB runs code
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.
- 11d ago First seen · 205 lines · 46 tokens per session scan A 4745aa670864
yq-windows-trash-cleaner is a skill published in the GitHub repository EthanYoQ/Skill-hub (9 stars, last pushed 5d ago), licensed MIT. It adds 46 tokens to every session and 3,051 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 2026-08-31.
Other skills, from other repositories
superplan-entry
Use when a new request might involve repo work, tracked changes, workspace-specific decisions, or durable coordination and you must decide whether Superplan should engage before exploring or implementing.
superplan-shape
Use when Superplan has decided to engage and the work still needs plans, specs, proof paths, or multiple tracked tasks shaped before bounded execution can start.
superplan-execute
Use when tracked work is already shaped and the next move is to execute, verify, block, or request feedback without replanning from scratch.
superplan-route
Use when Superplan may engage and you need to decide whether to stay lightweight, create one tracked task, gather context first, or shape deeper structure before execution begins.
superplan-brainstorm
Use when routed repo work still has meaningful ambiguity about behavior, constraints, taste, or acceptance intent, and that ambiguity would make planning or execution unsafe.
superplan-review
Use when a tracked task appears done and you need to decide whether its acceptance criteria are truly satisfied by fresh evidence.