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 skills/closetheloops/moradin/reviewnpx skills add CloseTheLoops/moradin --skill reviewgit clone --depth 1 https://github.com/CloseTheLoops/moradinWrote 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/closetheloops/moradin/review)<a href="https://agentmods.dev/skills/closetheloops/moradin/review"><img src="https://agentmods.dev/badge/skills/closetheloops/moradin/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 | $0.00061 | $0.00654 |
| Opus 5 | $0.00030 | $0.00327 |
| Sonnet 5 | $0.00012 | $0.00131 |
| Haiku 4.5 | $0.00006 | $0.00065 |
Grade A, and why
moradin-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 — 33 lines — stays where its author put it; the contents beside it link to each section on GitHub.
moradin:review
The stage that makes the Success Contract mean something. A kill line nobody reads is decoration; this stage reads it.
When invoked
/moradin:review — the dispatcher offers it when .forge/measure.md's review date has passed, when the contract's sample size is reached, or whenever the operator returns asking "how's it going?" Reads measure.md, backlog.md, journal.md.
Step 1 — Pull the actuals
From the instrumentation the build wired (the analytics events, the outcome log, the stats query — whatever the contract's "measured by" column names). Present them against the contract, side by side, in plain language: metric, target, actual, kill line, sample size so far. If the sample is still too small for the kill line, say so and schedule the next look — don't judge early.
Step 2 — Route, explicitly
- Continue — at or trending toward target: pick the next milestone from the backlog (statuses
idea/deferredare the menu), re-enter the pipeline at build (or research, if the milestone needs new decisions). - Pivot — kill line hit but the problem still looks real: the idea changes, not just the code. Route through the dispatcher's amendment into define; supersede the affected plan sections and decisions; write a lesson about what the numbers taught.
- Stop — kill line hit and the thesis is dead: close the project cleanly (run
/moradin:closeCore), capture the lesson, mark the journalstopped: <why>. Say it straight: a clean kill is the process succeeding — two months not spent building the wrong thing is the win the contract bought.
The routing is the operator's call — present the numbers and a recommendation, then wait. Never soften a kill-line breach into "let's keep tweaking"; that's the failure mode the contract exists to prevent.
Step 3 — Refresh the contract
Whatever the route, the contract updates: new stage label if the product graduated (validation → launch → growth), metrics re-drawn for the new stage (yesterday's north star is tomorrow's vanity metric), fresh targets, fresh review date (+30d), and the anti-vanity list re-checked. Supersede sections, don't silently edit — the old contract is the record of what we believed.
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 · 33 lines · 0 tokens per session scan A e24b34066738
moradin-review is a skill published in the GitHub repository CloseTheLoops/moradin (2 stars, last pushed 27d ago), licensed MIT. It adds 61 tokens to every session and 654 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
loop-engineering
Use when a repeatable task must become a bounded Trigger -> Execute -> Verify -> State loop, scheduled automation, goal agent, or metric-driven research cycle.
anthropic-os
Use when a personal or team operating system needs a bounded redesign using Four-C, closed-loop controls, 70/30 allocation, 3B creativity, experiments, and prediction-error learning.
knowledge-ops
Use when an Obsidian knowledge system needs classification, deduplication, retrieval, synchronization, debt queues, or governed Agent/Wiki promotion.
verify-before-claim
Use when an agent is about to claim completion, correctness, safety, publication, deployment, or any consequential external fact.
cognitive-compile
Use when source material must be transformed into a compact, evidence-aware model for learning, decisions, or an Obsidian concept note.
creativity-engine
Use when a defined problem needs diverse ideas, cross-domain combinations, and cheap experiments instead of a single untested answer.