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.
git 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/commands/coltonbearden/carrel/dedupe)<a href="https://agentmods.dev/commands/coltonbearden/carrel/dedupe"><img src="https://agentmods.dev/badge/commands/coltonbearden/carrel/dedupe/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/commands/coltonbearden/carrel/dedupe"><img src="https://agentmods.dev/badge/commands/coltonbearden/carrel/dedupe.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 | $0.00024 | $0.00621 |
| Opus 5.5 | $0.00010 | $0.00248 |
| Sonnet 5.5 | $0.00005 | $0.00124 |
| Haiku 4.5 | $0.00002 | $0.00062 |
Grade A, and why
dedupe 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 1mo 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.
What it actually says
Find duplicates in: $ARGUMENTS
Run the carrel CLI via Bash. Map the user's request onto the real flags in the --help block below (regenerated from the CLI by scripts/sync_plugins.py; if the installed carrel dedupe --help differs, trust the installed version — never invent flags):
Usage: carrel dedupe [OPTIONS] DIRS...
Report duplicate files under DIRS (recursively; hidden entries skipped).
Default is report-only. Deletion needs BOTH --delete newest|oldest AND --apply; without --apply
the deletions are only planned. The kept member of each group is never deleted. JSON output is a
list of {hash, files, kept, deleted}.
Options:
--near Perceptual matching for images (64-bit dHash, Hamming distance <= 8)
instead of exact content hashing. Non-image files are ignored in this
mode.
--delete [newest|oldest] Which duplicates to delete per group (by mtime); the other end of the
range is kept. Requires --apply to actually remove files.
--apply Actually delete (only together with --delete).
--json Machine-readable JSON output.
--help Show this message and exit.
Note: --json is a global flag and may come before the subcommand.
- Exact duplicates: content hash groups (size-prefiltered).
--near: images only, perceptual dHash clustering (catches resized/re-encoded copies). - Report-only is the default — nothing is ever deleted without BOTH
--delete <policy>AND--apply. Never pass--applyunless the user explicitly confirmed deletion after seeing the report; the kept member of each group is never deleted.
If the installed carrel predates dedupe (its --help reports no such command), say so; you may fall back to comparing carrel inspect --json sha256 values for a small set of files, but do not guess flags.
Interpret the JSON [{hash, files, kept, deleted}]: show each duplicate group, which copy would be kept, and the reclaimable bytes. Then ask before any destructive re-run.
Requires the carrel CLI on PATH. If carrel is not found, tell the user to install it with uv tool install carrel (see the repo's INSTALL notes), or run it as uv run carrel ... from the carrel repo root.
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.
- 1mo ago Changed · +19 lines 7dcbb27fd777
- 1mo ago Changed 07d03205f3ec
- 1mo ago First seen · 25 lines · 24 tokens per session scan A b1edeb427fbb
dedupe is a command published in the GitHub repository coltonbearden/carrel (2 stars, last pushed yesterday), licensed MIT. It adds 24 tokens to every session and 621 once invoked, about $0.0001 per session on Opus 5.5. 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 commands, from other repositories
review-resume
Comprehensive PM resume review against 10 best practices — structure, impact metrics, keywords, and actionable feedback.
follow-up
Review and act on pending reminders. Bare invocation shows all reminders as an indexed list; cron-fired invocation (/follow-up due ) surfaces the specific reminder and waits for a response; /follow-up review lists stale project follow-up entries for per-item disposition (extend/close/promote/skip). Transport-aware…
stop
Gracefully close the Vibe Squad tmux session from inside Chrono.
amp-download
Command "amp-download" from agentmessaging/claude-plugin, covering /amp-download, usage, arguments, options and examples.
todo.template
This prompt was authored for Claude-style slash workflows. In Codex runtime, adapt tool calls as follows.
maestro-companion
Quick execution for small tasks — minimal run lifecycle (start + done) with evidence recording. Full LLM capability, scoped to mechanically clear tasks.