unity-learn

unity-learn is a command for Claude Code from XeldarAlz/everything-claude-unity. It costs 22 tokens per session (1,434 once invoked), scanned A, original, MIT.

A Unity command for reviewing notes collected from earlier sessions and turning repeated patterns into draft skills. It reads the project’s stored learning records.

In plain words
What is it for?
Use it to summarize session history, inspect file-activity patterns, and identify possible new skills for the project.
Why use it?
It makes recurring lessons easier to find and reuse instead of leaving them buried in past session data.

Command for Claude Code

Part of the everything-claude-unity plugin — 10 skills, 27 commands, 20 agents, 5 hooks shipped together

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.

agentmods
npx agentmods add commands/xeldaralz/everything-claude-unity/unity-learn
Clone the repo
git clone --depth 1 https://github.com/XeldarAlz/everything-claude-unity

Made for: Claude Code.

Or install everything-claude-unity, the plugin that ships this one along with the rest of its 10 skills, 27 commands, 20 agents, 5 hooks.

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 unity-learn

README.md
[![agentmods](https://agentmods.dev/badge/commands/xeldaralz/everything-claude-unity/unity-learn.svg)](https://agentmods.dev/commands/xeldaralz/everything-claude-unity/unity-learn)
Your own site
<a href="https://agentmods.dev/commands/xeldaralz/everything-claude-unity/unity-learn"><img src="https://agentmods.dev/badge/commands/xeldaralz/everything-claude-unity/unity-learn.svg" alt="Measured on agentmods" height="20"></a>
Per session 22 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,434 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00022 $0.01434
Opus 5 $0.00011 $0.00717
Sonnet 5 $0.00004 $0.00287
Haiku 4.5 $0.00002 $0.00143

Measured 3d ago against content hash 7ec4b1791d32, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

unity-learn 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 3d 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.

.claude/commands/unity-learn.md · 167 lines

How it starts

The opening of the file, as written. The whole thing — 167 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/unity-learn — Learning Pipeline

Manage and leverage accumulated session learnings: $ARGUMENTS

This command works with the data collected by the auto-learn.sh hook (strict profile) which records session patterns to .claude/state/learnings.jsonl after each session. For pre-v1.3.0 projects, the file may be at .claude/learnings.jsonl instead.

Subcommands

review (default)

Read .claude/state/learnings.jsonl and present a dashboard summarizing accumulated data:

  1. Read the learnings file at .claude/state/learnings.jsonl (or .claude/learnings.jsonl as fallback for pre-v1.3.0 projects)
  2. Aggregate and present:
## Session Learning Dashboard

**Total sessions:** [count]
**Date range:** [earliest] to [latest]
**Total duration:** [hours]h [minutes]m

### File Activity
| Category | Total Edits | Sessions |
|----------|-------------|----------|
| Models   | [count]     | [count]  |
| Views    | [count]     | [count]  |
| Systems  | [count]     | [count]  |
| Tests    | [count]     | [count]  |
| Shaders  | [count]     | [count]  |
| Editor   | [count]     | [count]  |

### Session Categories
| Category     | Count | Avg Duration |
|-------------|-------|--------------|
| bug-fix     | [n]   | [m]m         |
| performance | [n]   | [m]m         |
| architecture| [n]   | [m]m         |
| workflow    | [n]   | [m]m         |
| integration | [n]   | [m]m         |

### Tool Usage
| Tool  | Total Calls | Avg per Session |
|-------|-------------|-----------------|
| Edit  | [count]     | [avg]           |
| Read  | [count]     | [avg]           |
| Bash  | [count]     | [avg]           |
| ...   | ...         | ...             |

extract

Analyze the learnings log for recurring patterns and apply confidence scoring:

  1. Read all entries from .claude/state/learnings.jsonl (or .claude/learnings.jsonl as fallback)
  2. Group by category (bug-fix, performance, architecture, workflow, integration)
  3. Identify recurring patterns:
    • Files that appear across multiple sessions → likely hotspots
    • Categories that dominate → project's current focus area
    • Tool usage patterns → workflow optimization opportunities
    • MVS balance → are Models, Views, and Systems equally maintained?
  4. Apply confidence scoring:
    • High confidence (3+ sessions): Pattern is well-established, likely a real project convention
    • Medium confidence (2 sessions): Pattern is emerging, worth noting but may be coincidental
    • Low confidence (1 session): Single observation, keep but don't act on yet
  5. Present findings:

Read the full file on GitHub · 167 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. 3d ago First seen · 167 lines · 22 tokens per session scan A 7ec4b1791d32

Subscribe to this mod's changes

unity-learn is a command published in the GitHub repository XeldarAlz/everything-claude-unity (21 stars, last pushed 4mo ago), licensed MIT. It adds 22 tokens to every session and 1,434 once invoked, about $0.0001 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 2026-08-30.