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 coltonbearden/carrel --skill watch-automationgit clone --depth 1 https://github.com/coltonbearden/carrelWrote 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/coltonbearden/carrel/watch-automation)<a href="https://agentmods.dev/skills/coltonbearden/carrel/watch-automation"><img src="https://agentmods.dev/badge/skills/coltonbearden/carrel/watch-automation/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/coltonbearden/carrel/watch-automation"><img src="https://agentmods.dev/badge/skills/coltonbearden/carrel/watch-automation.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 | $0.00053 | $0.00983 |
| Opus 5 | $0.00026 | $0.00491 |
| Sonnet 5 | $0.00011 | $0.00197 |
| Haiku 4.5 | $0.00005 | $0.00098 |
Grade A, and why
watch-automation 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 today.
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 — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Folder automation recipes with carrel watch
carrel watch DIR --on EVENTS --run CMD runs shell actions on file events (watchdog-based). {path} in the action is substituted with the triggering file, already shell-quoted. Multiple --run flags execute in order. --debounce MS collapses editor save-storms. Use --once or --timeout SECS for bounded test runs, --json-lines for machine-readable logs. Ctrl-C exits cleanly.
Before composing a recipe, run carrel watch --help to confirm the installed version has the flags you plan to use.
Recipe: auto-thumbnail new images
carrel watch ~/Pictures/incoming --on created --glob '*.png' \
--run 'carrel thumb {path} --out-dir ~/Pictures/thumbs'
Add a second --glob-less watch or broaden the glob for jpg. Thumbnails land in --out-dir; the watch ignores files its own actions write.
Recipe: auto-index documents into the desk db
Keep carrel search results fresh as files arrive:
carrel watch ~/Documents/desk --on created,modified \
--run 'carrel --root ~/Documents/desk index --update {path} --if-indexed'
--update (re)indexes just the touched file; --if-indexed makes it a silent no-op until someone has run carrel index once — safe to leave running.
Recipe: auto-convert a drop folder
Everything dropped as markdown comes out as PDF:
carrel watch ~/dropbox/md-in --on created --glob '*.md' \
--run 'carrel convert {path} --to pdf --out-dir ~/dropbox/pdf-out'
Chain steps with repeated --run (they execute sequentially per event), e.g. convert then index the output directory.
Recipe: file an inbox instead of scripting it
When the actions would be "read the document, name it, move it, index it", use carrel intake — one command instead of a --run chain, and it never overwrites or deletes:
carrel intake ~/inbox --to ~/archive # dry-run: the plan
carrel intake ~/inbox --to ~/archive --apply # file what is there
carrel intake ~/inbox --to ~/archive --watch --stable 5 # keep filing arrivals
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.
- today Changed · +22 lines 50fd321e53df
- 11d ago First seen · 49 lines · 53 tokens per session scan A 91fc05af32cd
watch-automation is a skill published in the GitHub repository coltonbearden/carrel (1 stars, last pushed today), licensed MIT. It adds 53 tokens to every session and 983 once invoked, about $0.0003 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
ade
Parse documents and extract schema-shaped data with the ADE (Agentic Document Extraction) v2 APIs through the ade CLI. A local job-item store makes every run idempotent, resumable, and citable — repeat runs are free, interrupted runs resume, and every answer can cite element ids with visual evidence.
document-processing
Use when the deliverable is a document's bytes or its literal content — text/tables out of PDFs, AcroForm fill and flatten, page merge/split, PDF/DOCX from templates, OCR of image-only scans. NOT schema-typed fields pulled from text (that is structured-extraction), NOT signature routing (e-signature) or spreadsheet…
c-screen
Capture screenshots and extract text via OCR using peekaboo, and capture webcam images using camsnap. Enables visual analysis of screen content and camera input.
quality-gates
This skill should be used when the user wants to run code quality checks (linting, formatting, type checking, tests) on the lazyclaude project. Use this skill when asked to "run quality gates", "check the code", "run tests", "lint the code", or verify code quality before committing.
update-marketplace-stars
Update GitHub star counts for suggested marketplaces in DEFAULTSUGGESTEDMARKETPLACES. Use this skill when asked to "update stars", "refresh marketplace stars", or "sync star counts".
kuake_skill
Enables OpenClaw agents to interact with Quark Cloud Drive using the kuake CLI tool for file operations like listing, uploading, downloading, and sharing.