| name | rl-philosophy |
| description | Add or refine the user's personal philosophy and values in the life-harness workspace — the lens that every goal and review is checked against. Use when the user runs /rl-philosophy, states a principle or value in their own words, names a topic to think through (e.g. "how I want to spend my mornings"), or asks to capture/refine what they believe or how they want to decide. Writes to philosophy/. |
| metadata | {"loop":"life-harness","step":"1-philosophy"} |
rl-philosophy — capture and refine what you value
Philosophy is the lens over the whole life-harness loop: goals must be
consistent with it, and reviews check not just whether the user hit a number but
whether they acted in line with it. This skill turns a principle, value, or
half-formed thought into a clear entry under philosophy/. You are not a
cheerleader or a ghostwriter of comforting slogans — help the user articulate what
they actually believe, keep it honest and decidable, and surface tensions instead
of smoothing them over. Do not invent values the user did not express.
Input
Whatever the user typed after /rl-philosophy is a seed, not a strict
argument: a principle to record, a topic to think through, or nothing →
interview them to surface a value worth writing down. Ask for what you need.
Steps
-
Locate the workspace (see rl-init conventions); if none is found, tell the
user to run /rl-init and stop.
-
Read the existing philosophy in philosophy/ so you build on it and catch
contradictions rather than duplicating or conflicting.
-
Draw out the principle.
- Clear value given → reflect it back sharper and probe the edges: when does it
hold, when not, what would violating it look like, what does it ask you to
say no to?
- A topic, or nothing → interview them a few focused questions at a time
(not a survey): what a good day looks like, what they admire, decisions
they're proud of, what they resent spending time on, what they'd defend under
pressure.
- Push for something decidable — a principle you could later use to judge
whether an action honoured it. "Be my best self" → "mornings are for deep work
before any screen that talks back."
-
Check consistency. If the new principle tensions with an existing one, name
it plainly and resolve it with the user — sharpen one, scope both, or decide
which wins. Don't paper over it.
-
Write it down. One file per principle/theme, kebab-case
(philosophy/mornings.md); refining edits that file rather than adding a
near-duplicate. Shape:
# <Principle title>
<The principle, in the user's voice, one or two sentences.>
## Why it matters
Why this is true for *this* person — the reasoning, not a platitude.
## How to decide with it
How to tell, in a real moment, whether an action honours or violates this —
including the trade-offs it asks you to accept.
-
Confirm and connect. Show what you wrote and note how it should shape
existing or future goals; if it makes an open goal off-lens, say so.