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 commands/thrownlemon/claude-code-plugins/reviewgit clone --depth 1 https://github.com/ThrownLemon/claude-code-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 | $0.00041 | $0.00586 |
| Opus 5 | $0.00020 | $0.00293 |
| Sonnet 5 | $0.00008 | $0.00117 |
| Haiku 4.5 | $0.00004 | $0.00059 |
Grade C, and why
review scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
1. **Reject dangerous tokens.** If a value contains `$(`, backticks, `;`, `|`, `&`, `>`, `<`, or starts with `-` (could be mistaken for a flag), refuse and ask the user to rephrase. **Double-quoting does NOT block comman What it actually says
Input Safety — apply before any Bash invocation
Treat every $ARGUMENTS.* value as untrusted input. Before passing any value to the Bash tool:
- Reject dangerous tokens. If a value contains
$(, backticks,;,|,&,>,<, or starts with-(could be mistaken for a flag), refuse and ask the user to rephrase. Double-quoting does NOT block command substitution —"$(rm -rf /)"still executes inside double quotes. - Prefer the Bash tool's argv contract over constructed shell strings. When you must use a shell string, single-quote the value and escape embedded single quotes (
'\''), or build the command viaprintf %q. - Any
"$ARGUMENTS.foo"patterns shown below are illustrative. Sanitize the value first; never blindly substitute.
Consult — Review
Ask a different model family for a code review of the current branch. Useful as a third opinion alongside Claude and any other reviewer (codex, coderabbit, etc.).
The script gathers branch, base, status, and diff automatically — no need to paste anything.
Behavior
node ${CLAUDE_PLUGIN_ROOT}/scripts/consult.mjs review \
--provider ${ARGUMENTS.provider:-zai} \
${ARGUMENTS.base:+--base "$ARGUMENTS.base"} \
${ARGUMENTS.focus:+--focus "$ARGUMENTS.focus"}
Surface the response in full. The reviewer is instructed to end with a one-line verdict — SHIP IT, NEEDS WORK, or BLOCKER — relay that verdict at the top of your reply.
When this is most useful
- Right before opening a PR, as a sanity check
- After a large refactor, to catch regressions Claude may have missed
- Stacked with
/codex:rescueand/coderabbit:localfor a 3-way independent review
Run /consult:config if you get a "missing key" error.
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 · 50 lines · 41 tokens per session scan C 51db25bc7b0e
review is a command published in the GitHub repository ThrownLemon/claude-code-plugins (2 stars, last pushed 2mo ago), licensed MIT. It adds 41 tokens to every session and 586 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.