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/everywan-dev/claude-code-engineering/route-work-to-the-right-modelnpx skills add everywan-dev/claude-code-engineering --skill route-work-to-the-right-modelgit clone --depth 1 https://github.com/everywan-dev/claude-code-engineeringWrote 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/everywan-dev/claude-code-engineering/route-work-to-the-right-model)<a href="https://agentmods.dev/skills/everywan-dev/claude-code-engineering/route-work-to-the-right-model"><img src="https://agentmods.dev/badge/skills/everywan-dev/claude-code-engineering/route-work-to-the-right-model.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 | $0.00049 | $0.00964 |
| Opus 5 | $0.00024 | $0.00482 |
| Sonnet 5 | $0.00010 | $0.00193 |
| Haiku 4.5 | $0.00005 | $0.00096 |
Grade A, and why
route-work-to-the-right-model scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Route work to the right model
Deciding how much validation a change needs, per change, in the moment, is how teams end up validating nothing.
The decision always gets made under pressure, by the person who most wants the answer to be "not much". So it is made in advance, from the kind of work, by a table you can read and argue with when nothing is on fire.
Ask it
validated-memory route "what you are about to do" [--path FILE ...] [--json]
Running from the installed plugin, where the CLI is not on PATH:
PYTHONPATH="${CLAUDE_PLUGIN_ROOT}" python3 -m validated_memory route "what you are about to do"
It answers with a level, the number of validations, which agents, which model, and which effort. It exits 0 always: this advises, it does not gate.
$ validated-memory route "small tweak to the checkout flow" --path src/billing/refund.py
Level 3 — 3 validations
matched money: billing, checkout, refund
matched touched path: billing
model : opus
effort : high
agents : security-reviewer, data-reviewer, devils-advocate
Three, and the devil's advocate has to actively try to break it.
"Small tweak" is a description of intent. The paths and the words are a description of risk, and only one of those two decides.
The three levels
| Level | Kind of work | Validations | Model | Effort |
|---|---|---|---|---|
| 1 | Cosmetic, documentation, anything that never reaches production | 1 | small | low |
| 2 | Production code, service config, CI, dependencies | 2 independent | mid | medium |
| 3 | Auth, permissions, money, customer data, migrations, network, certificates, deletions | 3, one trying to break it | large | high |
The one hard rule
Nothing at level 3 runs on the small model.
Saving tokens is worth it exactly where being wrong is cheap. It is never worth
it where being wrong means an outage, a breach, or someone's money. That rule is
asserted in code, not left to good intentions — see
route.py and the tests that pair every
level-3 phrase against the model it was given.
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.
- 5d ago First seen · 99 lines · 49 tokens per session scan A 8cc2b5ebd3ca
route-work-to-the-right-model is a skill published in the GitHub repository everywan-dev/claude-code-engineering (2 stars, last pushed 4d ago), licensed Apache-2.0. It adds 49 tokens to every session and 964 once invoked, about $0.0002 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
claude-project-manifest
Creates annotated bibliography-style manifests for Claude projects, tracking files, conversation threads, and relationships with unique IDs and annotations. Use when documenting project contents, creating file inventories, tracking conversation history, or building navigable knowledge maps.
proofpress
Preserve or inspect decision history for Markdown and static HTML knowledge artifacts with Proofpress. Use when creating or accepting a meaningful revision to an ADR, design doc, PRD, README, strategy note, portable .md artifact, or static .html/.htm artifact in a configured repository, or when receiving a file…
book-to-skill
Use when convert technical books and documents (PDF, EPUB, DOCX, HTML, Markdown, RTF, MOBI) into structured agent skills with frameworks, mental models, chapter references, and decision rules. Includes full extraction pipeline. Use when the user wants to turn a book or document collection into a reusable agent skill…
proofpress-governed-context
Use Proofpress governed context when a task must rely on or hand off durable agent-produced knowledge, including evidence-backed decisions, experiment results, integration contracts, or incident learnings. Retrieve eligible context before relying on it; submit bounded proposals and receipts when a durable outcome is…
llm-wiki
Maintain a personal team knowledge base using the LLM Wiki pattern — incremental ingest, query, and lint operations on a layered wiki architecture.
Principle Synthesizer
Synthesize invariant principles from 3+ sources — find the core that survives across all expressions.