lean-learn

A command that checks a project hypothesis against observations and records what to do next. A hypothesis is an idea that still needs evidence; the possible decisions are continue, adjust, change direction, or stop.

In plain words
What is it for?
Use it to review a Lean-Track experiment, compare evidence with the assumptions in FRAME.md, and create a LEARNINGS.md file.
Why use it?
It turns scattered feedback and observations into a documented decision. This helps prevent continuing with an idea that the available evidence does not support.

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/jonase47/ccpr/lean-learn
Clone the repo
git clone --depth 1 https://github.com/jonase47/ccpr
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,644 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 $0.00000 $0.01644
Opus 5 $0.00000 $0.00822
Sonnet 5 $0.00000 $0.00329
Haiku 4.5 $0.00000 $0.00164

Measured yesterday against content hash fc01acc9ccef, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

lean-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/lean-learn.md · 144 lines

How it starts

The opening of the file, as written. The whole thing — 144 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/lean-learn – Validate Lean-Track and make a decision

Creates docs/LEARNINGS.md with hypothesis check, observations, and decision (PROMOTE / PIVOT-Soft / PIVOT-Hard / DROP). Closes a Lean cycle and determines what happens next.

Argument: $ARGUMENTS = [optional: projectdir]

  • Without argument: operates on $(pwd).

Prerequisites

  • docs/FRAME.md exists with hypothesis
  • Code exists in the repo (otherwise there is nothing to learn from)
  • Template: ~/.claude/templates/LEARNINGS_TEMPLATE.md

Lead Agent

konzeptor for hypothesis analysis, business-analyst as support for market/user signals.

Execution

1. Prerequisites check

  • If docs/FRAME.md is missing → STOP, note: run /lean-frame first
  • If no code files exist in src/ (or repo root) → warn, continue with user confirmation

2. Load hypothesis from FRAME

Read docs/FRAME.md section 2 (hypothesis) and section 7 (risk assumptions: what must work without fail?). List them for the user.

3. Walk through validation interactively

Ask interactively:

  1. Hypothesis check: Is the hypothesis confirmed, refuted, or unclear?

    • Evidence: What concrete data points/observations/user reactions are available? (Not "felt", but concrete.)
  2. What works? (factual observations, each with a brief rationale)

  3. What does not work? (observations with suspected cause)

  4. Surprises: What was unexpected? (If nothing: record explicitly — the hypothesis was probably too weak.)

4. Make a decision

Present 4 options:

Option When appropriate
PROMOTE Hypothesis confirmed; prototype becomes the basis for Full-Track
PIVOT-Soft New hypothesis on the same tech base; most code stays
PIVOT-Hard Wrong core assumption (tech, market, user); code reset required
DROP Lean answered the question (negatively); project is frozen

5. On PIVOT (Soft or Hard): module table

Ask per module/directory:

  • keep — functionally correct, hypothesis-independent
  • refactor — logic OK, interface must change
  • rebuild — wrong core assumption, rewrite

Read the full file on GitHub · 144 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 · 144 lines · 0 tokens per session scan A fc01acc9ccef

Subscribe to this mod's changes

lean-learn is a command published in the GitHub repository jonase47/ccpr (1 stars, last pushed yesterday), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,644 tokens. 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.