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 CATWILLgh/MAINFRAME --skill mainframe-decision-reviewgit clone --depth 1 https://github.com/CATWILLgh/MAINFRAMEWrote 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/catwillgh/mainframe/mainframe-decision-review)<a href="https://agentmods.dev/skills/catwillgh/mainframe/mainframe-decision-review"><img src="https://agentmods.dev/badge/skills/catwillgh/mainframe/mainframe-decision-review.svg" alt="Measured on agentmods" 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 | $0.00058 | $0.01111 |
| Opus 5 | $0.00029 | $0.00556 |
| Sonnet 5 | $0.00012 | $0.00222 |
| Haiku 4.5 | $0.00006 | $0.00111 |
Grade A, and why
mainframe-decision-review 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 6d 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 — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Decision review
Test the proposed decision against its strongest grounded failure modes. The goal is a more reliable decision, not disagreement for its own sake.
Grounding rule
Ground every material objection in repository evidence, current primary documentation, a bounded experiment, or explicit reasoning. Mark reasoning-only claims as the weakest evidence. If an objection cannot be grounded, discard it and name the limit instead of manufacturing doubt.
Treat a clean review as useful evidence. proceed is the correct verdict when
no serious grounded objection survives inspection.
Review method
- Reduce the brief to a neutral proposal, its boundary, constraints, viable alternatives, load-bearing assumptions, evidence, and cost of failure. Ignore confidence, enthusiasm, status, and claims that the answer is obvious.
- Inspect the affected code, configuration, contracts, and dependency chain before objecting. A failure already prevented by the real system is not a finding. Stay inside the supplied decision boundary.
- Verify every version-sensitive framework, library, protocol, API, security, or operational claim against current owning documentation. Use Context7 when it exposes the official corpus; otherwise use live web search to reach the primary source. Add an independent source only when the claim is disputed, interpretive, plausibly stale, or expensive to get wrong.
- Run a pre-mortem at a realistic failure horizon: assume the decision failed materially, work backwards to specific causes, then keep only causes that survive grounding.
- Try to falsify each load-bearing assumption. State what observation would make it false and whether that observation is present, absent, or still unavailable.
- Build the strongest alternative only for a genuine decision fork. Compare it under the same constraints and state where it wins and loses. Do not invent an alternative to fill the report.
- Rank surviving objections by consequence and evidence confidence. Stop when the load-bearing assumptions and strongest grounded objections are resolved; more searching is not automatically more confidence.
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.
- 6d ago First seen · 101 lines · 58 tokens per session scan A 6a230dbd40a1
mainframe-decision-review is a skill published in the GitHub repository CATWILLgh/MAINFRAME (2 stars, last pushed 16d ago), licensed MIT. It adds 58 tokens to every session and 1,111 once invoked, about $0.0003 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 skills, from other repositories
at-review
Review code changes for bugs, regressions, convention violations, and high-value cleanup opportunities. Use for diffs, commit ranges, hosted PR/MR URLs, branches, paths, staged changes, or working-tree changes.
pr-triage
4-phase PR backlog management with audit, deep code review, validated comments, and optional worktree setup. Use when triaging pull requests, catching up on pending code reviews, or managing a backlog of open PRs. Args: 'all' to review all, PR numbers to focus (e.g. '42 57'), 'en'/'fr' for language, no arg = audit…
audit-agents-skills
Audit Claude Code agents, skills, and commands for quality and production readiness. Use when evaluating skill quality, checking production readiness scores, or comparing agents against best-practice templates.
review-pr
Perform a comprehensive code review of a pull request.
validate-changes
Evaluate staged changes using LLM-as-a-Judge before committing.
review-plan
Structured plan review across 4 axes before writing any code (inspired by Garry Tan's workflow).