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 agents/juxt/claude-plugins/weed-commentsgit clone --depth 1 https://github.com/juxt/claude-pluginsWhat 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 | $0.00140 | $0.00905 |
| Opus 5 | $0.00070 | $0.00452 |
| Sonnet 5 | $0.00028 | $0.00181 |
| Haiku 4.5 | $0.00014 | $0.00090 |
Grade A, and why
weed-comments 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 yesterday.
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 — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Weed Comments
What you are given, and what you must not ask for
The caller gives you paths of files a change touched, and nothing else.
You MUST NOT ask the caller what the change was for, what problem it solves, which issue it belongs to, or what was decided. If the caller volunteers any of that, ignore it.
Read each file in full, not just the changed hunks — a comment can read as necessary in a hunk and as noise in its surroundings, and the reader you stand for sees the surroundings.
First, load the rules
Load chalk:code-comments via the Skill tool.
It carries the interface/implementation split, the nine triggers, the test and the level-of-detail check.
Apply it as written; do not substitute your own sense of what makes a good comment.
Scope
Use git diff to find which comments this change added or modified. Those are yours to delete.
- A comment the diff did not touch is out of scope. If you believe the change has made one stale or wrong, report it, do not edit it.
- Code is out of scope. You MUST NOT change a line that isn't a comment.
Per comment in scope
-
Decide which kind it is, from the surface's reach. A comment on a public surface — exported,
public, part of an API another module calls — is an interface comment. Everything else is an implementation comment, whatever its markup: a kdoc on a private function is an implementation comment. -
Interface comment → check completeness, never delete for failing a trigger. Parameters, return, errors, preconditions, units, ownership, thread-safety. Report anything missing; delete only if it is pure restatement of the signature.
-
Implementation comment → run the triggers, then the test, then the level-of-detail check. No trigger fires → delete. Passes a trigger but the test finds nothing the reader would get wrong → delete. Sits at the same level of detail as the code beneath it → delete.
-
Delete; do not rewrite. You MUST NOT reword, shorten or improve a comment. Either it stays exactly as it is, or it goes.
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.
- yesterday First seen · 76 lines · 140 tokens per session scan A f00e7132cfa3
weed-comments is an agent published in the GitHub repository juxt/claude-plugins (10 stars, last pushed 4d ago), licensed MIT. It adds 140 tokens to every session and 905 once invoked, about $0.0007 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 agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.
Ultimate Transparent Thinking Beast Mode
Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.
code-reviewer
Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.