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 mehrad-dm/mastermind --skill deepengit clone --depth 1 https://github.com/mehrad-dm/mastermindWrote 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/mehrad-dm/mastermind/deepen)<a href="https://agentmods.dev/skills/mehrad-dm/mastermind/deepen"><img src="https://agentmods.dev/badge/skills/mehrad-dm/mastermind/deepen/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/mehrad-dm/mastermind/deepen"><img src="https://agentmods.dev/badge/skills/mehrad-dm/mastermind/deepen.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00108 | $0.01025 |
| Opus 5.5 | $0.00043 | $0.00410 |
| Sonnet 5.5 | $0.00022 | $0.00205 |
| Haiku 4.5 | $0.00011 | $0.00103 |
Grade A, and why
deepen 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 15d 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 — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MasterMind: Deepen
Find where the design costs the most, present the candidates, let the user choose. You survey and
rank. You do not restructure: the chosen one goes to the refactorer agent, which is where the edits
happen and where behaviour is proven unchanged.
Load ~/.mastermind/engineering/core/principles.md and mindset.md for the vocabulary. Read the
field's audit-rules.md and .mastermind/brief.md if they exist: with no field pack you still
have the universal principles, and with no brief you use the codebase's own names.
Scope it before you read anything
A whole-repo survey produces a list nobody acts on. Narrow first:
- Take the area the user named. If they named none, use
git logto find what changed most in the last few months. Design debt hurts where the code moves. - Cap it. Six candidates maximum, and fewer is better.
- Send the reading to a subagent where you have one, so the survey does not eat this context. Where you do not, read in the same order and keep notes short: the cap is what protects the context.
What you are looking for
Not ugliness. Cost. Each of these has a named future price:
- A shallow module: the interface costs about as much to understand as the implementation it hides. It adds a name and buys nothing.
- A concept spread thin: understanding one idea means opening five files. The split was mechanical, not conceptual.
- A leaked boundary: callers know the shape of what is behind the seam, so the seam cannot move.
- A path that cannot be tested without standing up the world, because there is nowhere to substitute.
- An invariant enforced in several places, which will drift, because one day someone edits four of five.
Apply the deletion test to anything suspected shallow: if this module disappeared and its callers did the work inline, would the code get worse? If the honest answer is no, that is the finding.
Present them as a choice, not a report
One block per candidate, ranked, in plain text. Use the forms in
~/.mastermind/engineering/core/clarity.md: a tree, a call flow, a before and after. Plain text over a
rendered diagram, because it reads faster and works in every tool.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 15d ago Changed · +108 tokens per session b2f022f24fe9
- 1mo ago First seen · 77 lines · 0 tokens per session scan A 2a3ed68e9458
deepen is a skill published in the GitHub repository mehrad-dm/mastermind (23 stars, last pushed 6d ago), licensed MIT. It adds 108 tokens to every session and 1,025 once invoked, about $0.0004 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 2026-08-30.
Other skills, from other repositories
code-review
Run a thorough self-review pass on the most recent change.
code-review
Use when reviewing code for bugs, security issues, performance problems, or adherence to best practices. Provides structured code review with severity levels and actionable feedback.
code-coherence-review
Reviews code and plans for logic coherence, correctness, and structural integrity.
reviewer-architect-adversarial
Adversarial plan review — structural validation and assumption attack on grill artifacts before implementation begins.
code-quality-review
Reviews code and plans against the project's coding rules.
review-loop
Two-mode review loop — single reviewer per epic, three reviewers for full branch.