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 lacerbi/dotclaude --skill deepthinkgit clone --depth 1 https://github.com/lacerbi/dotclaudeWrote 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/lacerbi/dotclaude/deepthink)<a href="https://agentmods.dev/skills/lacerbi/dotclaude/deepthink"><img src="https://agentmods.dev/badge/skills/lacerbi/dotclaude/deepthink/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/lacerbi/dotclaude/deepthink"><img src="https://agentmods.dev/badge/skills/lacerbi/dotclaude/deepthink.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 | $0.00011 | $0.01262 |
| Opus 5.5 | $0.00004 | $0.00505 |
| Sonnet 5.5 | $0.00002 | $0.00252 |
| Haiku 4.5 | $0.00001 | $0.00126 |
Grade A, and why
deepthink 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.
How it starts
The opening of the file, as written. The whole thing — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Problem
$ARGUMENTS
Configuration
Defaults: 5 agents for Cycle 1, 3 agents for subsequent cycles, 2–3 cycles (2 if early convergence, 3 if significant disagreement persists). Adjust if the user specifies (e.g., "use 10 agents", "do 4 rounds", "go deeper").
Your Task
Achieve higher-quality analysis through parallel independent reasoning with iterative refinement.
You (main agent) orchestrate and synthesize. Sub-agents analyze based on the context you provide.
Phase 0: Context Preparation
Before spawning any sub-agents, ensure you have sufficient context.
If you lack understanding of the problem domain, codebase, or relevant materials:
- Explore the project structure
- Read relevant files in full
- Gather whatever context is needed to deeply understand the problem
If you already have sufficient context from the current conversation, proceed directly.
Critical: Sub-agents will only know what you tell them. They start cold with no access to your conversation history or accumulated understanding.
Spawning Sub-Agents: Context Requirements
When spawning each sub-agent, you must include in the prompt:
- Full problem statement with background and goals
- Relevant file contents (or key excerpts) that inform the analysis
- Key findings you've already established
- Specific constraints or requirements the user has mentioned
- The assigned lens for that agent's perspective
Do NOT assume sub-agents will figure out context from a thin prompt. Give them everything they need to reason well.
Sub-Agent Instructions
When spawning sub-agents, instruct them to:
- Be extremely thorough—consider the problem deeply until fully satisfied with their analysis
- Spend maximum effort; do not satisfice or stop at "good enough"
- Start from the provided context, then explore further to deepen understanding, fill gaps, and verify assumptions
- Only conclude when they have genuinely exhausted their reasoning on the problem
- Produce: analysis, conclusion, confidence (high/medium/low), key assumptions made
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 First seen · 138 lines · 11 tokens per session scan A 4703e656a843
deepthink is a skill published in the GitHub repository lacerbi/dotclaude (2 stars, last pushed yesterday), licensed MIT. It adds 11 tokens to every session and 1,262 once invoked, about $0.0000 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 skills, from other repositories
bet-on-it
Use when debugging or changing behavior based on an uncertain causal hypothesis that the next action can test.
prove-me-wrong
Use when a leading diagnosis or nontrivial fix has initial support and needs one adversarial counterexample before acceptance.
archaeologist
Use when changing unusual or defensive code, mature APIs, dependencies, or architecture, or estimating a migration where Git history may reveal hidden constraints.
no-vibes
Use when completion depends on an end-to-end outcome across components, environments, or external systems.
red-button
Use when an action could cause material, hard-to-reverse harm to production, data, security, finances, users, or external systems.
organizing-files
Organizes macOS files across Desktop, Documents, Downloads, and iCloud Drive into a consistent structure. Use when the user asks to organize files, clean up folders, sort downloads, declutter desktop, tidy up documents, or structure their filesystem. Triggers on "organize", "clean up", "sort files", "declutter", "file…