oxy:eval-and-improve

oxy:eval-and-improve is a command for Claude Code from oxy-hq/skills. It costs 35 tokens per session (2,384 once invoked), scanned B, original, MIT.

A command that evaluates Oxy skills against written scoring rules using a project's real database, then proposes and applies changes to their instruction files.

In plain words
What is it for?
Use it to test all four Oxy skills, or one selected skill, in a temporary project and update its SKILL.md instructions based on the results.
Why use it?
It helps find where the skills produce weak or incorrect results and turns those findings into targeted improvements. It also checks that the project is suitable for evaluation first.

Command for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: reads .claude/ paths; mentions CLAUDE.md.

Part of the oxy-skills plugin — 8 skills, 6 commands, 1 agent shipped together

Good fit Use it to test all four Oxy skills, or one selected skill, in a temporary project and update its SKILL.md instructions based on the results.

Compare 6 commands from other repositories ↓
Install with agentmods
npx agentmods add commands/oxy-hq/skills/eval-and-improve
Install

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.

Clone the repo
git clone --depth 1 https://github.com/oxy-hq/skills

Made for: Claude Code.

Or install oxy-skills, the plugin that ships this one along with the rest of its 8 skills, 6 commands, 1 agent.

Wrote 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.

agentmods badge for oxy:eval-and-improve

README.md
[![agentmods](https://agentmods.dev/badge/commands/oxy-hq/skills/eval-and-improve.svg)](https://agentmods.dev/commands/oxy-hq/skills/eval-and-improve)
Your own site
<a href="https://agentmods.dev/commands/oxy-hq/skills/eval-and-improve"><img src="https://agentmods.dev/badge/commands/oxy-hq/skills/eval-and-improve.svg" alt="Measured on agentmods" height="20"></a>
Per session 35 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,384 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 2 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00035 $0.02384
Opus 5 $0.00017 $0.01192
Sonnet 5 $0.00007 $0.00477
Haiku 4.5 $0.00003 $0.00238

Measured 7d ago against content hash 76edc3734b4c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade B, and why

oxy:eval-and-improve scanned grade B with 2 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 7d 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

grep -rl '"name": "oxy-skills"' ~/.claude/plugins ~/.claude/projects ~/Documents 2>/dev/null \

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -s https://raw.githubusercontent.com/oxy-hq/oxy/refs/heads/main/json-schemas/agent.json
commands/eval-and-improve.md · 240 lines

How it starts

The opening of the file, as written. The whole thing — 240 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Eval and Improve Oxy Skills

Evaluate all 4 oxy skills (or one specific skill) by running them against the current project's real database, scoring the output against rubrics, and proposing targeted improvements to each SKILL.md file.

Prerequisites Check

Before starting, verify:

  1. You are in a copy of a real oxy project, not the live repo. Run pwd and confirm this is a temp/eval directory, not production.

  2. Skill output directories are clean. The eval tests what each skill generates from scratch. Before starting, remove any pre-existing skill outputs so skills aren't judged on files they didn't create:

    rm -f semantics/views/*.view.yml semantics/topics/*.topic.yml
    rm -f workflows/*.workflow.yml *.agent.yml *.app.yml
    

    Keep: config.yml, semantics.yml, example_sql/, globals/, .env

  3. Database schema is synced: Run oxy sync if semantics.yml is missing or stale.

  4. oxy build requires a running PostgreSQL instance. Run oxy start --enterprise in a separate terminal before starting this eval session. oxy start --enterprise launches a Docker PostgreSQL container and the oxy web server. It exposes PostgreSQL on port 15432 with default credentials. Add this to the project's .env so oxy can connect:

    OXY_DATABASE_URL=postgresql://postgres:postgres@localhost:15432/oxy
    

    Note: oxy --version ≥ 0.5.27 is only needed for oxy run --dry-run on SQL files, not for oxy build.

  5. Export OXY_DATABASE_URL at the start of this session. Each Bash tool call runs in a fresh shell, so export does not persist between calls. Read the value from .env once and prepend it to every oxy command:

    export OXY_DATABASE_URL=$(grep OXY_DATABASE_URL .env | cut -d= -f2-)
    

    Then use this pattern for all oxy commands throughout the eval:

    OXY_DATABASE_URL=$(grep OXY_DATABASE_URL .env | cut -d= -f2-) oxy build
    OXY_DATABASE_URL=$(grep OXY_DATABASE_URL .env | cut -d= -f2-) oxy run ...
    

Read the full file on GitHub · 240 lines

Changes

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.

  1. 7d ago First seen · 240 lines · 35 tokens per session scan B 76edc3734b4c

Subscribe to this mod's changes

oxy:eval-and-improve is a command published in the GitHub repository oxy-hq/skills (2 stars, last pushed today), licensed MIT. It adds 35 tokens to every session and 2,384 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 2 findings (reads agent configuration directories, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.