app-learn

A command that studies a shipped software project and adds its conventions and failures to a shared House Knowledge Base.

In plain words
What is it for?
Use it after shipping an app, or on an existing app, to inspect its documentation and build files, import recorded learnings, update knowledge pages, and record failures.
Why use it?
It turns lessons discovered during development into reusable guidance and highlights conflicts between written learnings and the code that was actually shipped.

Command

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/vmobifystudio/app-dev-team/app-learn
Clone the repo
git clone --depth 1 https://github.com/vmobifystudio/app-dev-team
Per session 41 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,251 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 $0.00041 $0.02251
Opus 5 $0.00020 $0.01125
Sonnet 5 $0.00008 $0.00450
Haiku 4.5 $0.00004 $0.00225

Measured yesterday against content hash ae690162a848, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

app-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 yesterday.

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.

commands/app-learn.md · 157 lines

How it starts

The opening of the file, as written. The whole thing — 157 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/app-learn — Grow the House Knowledge Base

Paths to mine (default = the current/just-shipped project): $ARGUMENTS

The House KB (knowledge/) is living. After shipping an app — or to ingest an existing one — this command folds its real conventions back into the packs so future apps start smarter.

Steps

  1. Resolve targets. For each path, confirm it's a real app (has CLAUDE.md, README.md, docs/ARCHITECTURE.md, and/or build files). If none given, use the current project.

1a. Read the learnings inbox first — docs/90-learnings.md. /app-ship harvests every LEARNING: line the team wrote during the run into it, each with the daily fragment it came from. These are the highest-value inputs here: they are conventions the team discovered while building, not conventions inferred afterwards from finished code. Treat each as a candidate convention going into step 3's diff, alongside what step 2 mines from the app itself. Where a harvested line and the mined code disagree, the code wins and the line is recorded as a conflict — a learning nobody ended up following is not a house rule.

No inbox (an app mined from outside a /app-ship run) → say so and continue with step 2 alone.

  1. Mine in parallel. Spawn one general-purpose Agent per app (in a single message) to extract, read-only, a structured report: stack & versions, architecture, state, persistence, DI, navigation, networking, testing, monetization (IAP/ads/consent), analytics, ASO/store, git/commit conventions, and explicit "always/never" house rules. (This mirrors the original 7-app mining that seeded the KB.)

  2. Diff against the KB. For each pack under knowledge/, compute:

    • New conventions not yet captured → propose adding them.
    • Confirmations of existing rules → note increased confidence, no change.
    • Conflicts — the app disagrees with a pack (e.g. a different DI, analytics default, or min-SDK). Never silently overwrite. Record both positions and surface the conflict.

Read the full file on GitHub · 157 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. yesterday First seen · 157 lines · 41 tokens per session scan A ae690162a848

Subscribe to this mod's changes

app-learn is a command published in the GitHub repository vmobifystudio/app-dev-team (4 stars, last pushed 22d ago), licensed MIT. It adds 41 tokens to every session and 2,251 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.