Liftoscript Program Authoring

Output contract

Default to slash-style format for new lines:

Exercise Name / set expressions / section: value / section: value

Keep one separator style per line; do not mix slash and freeform section placement.
Preserve existing local style when editing unless the user asks to normalize.
Use explicit section names for non-set data (
```
progress:
```
,
```
update:
```
,
```
warmup:
```
,
```
superset:
```
,
```
used:
```
,
```
id:
```
).

Keep terminology consistent: use "exercise line", "set", "progress", "update", "state variable", "set variation".
Prefer built-in progression (
```
lp
```
,
```
dp
```
,
```
sum
```
) before custom scripts.
For scripting tasks, identify execution context first:
```
progress: custom()
```
vs
```
update: custom()
```
.
Use timer literals by context:
```
45s
```
/
```
120s
```
in exercise-line sets, numeric timers in scripts (
```
timers[5] = 20
```
).
Reject timed rep tokens in set syntax (
```
3x60s
```
); use numeric reps (
```
3x60
```
) when needed.
Resolve units deterministically: nearby program text -> explicit user preference -> default
```
kg
```
.
Do not silently mix
```
kg
```
and
```
lb
```
in one generated snippet.
In exercise-line set syntax,
```
%
```
load tokens are percentages of
```
1RM
```
(not training max).

Resolve exercise naming from

references/exercise-name-resolution.md

against

references/exercise-list.md

In full mode, place template lines (
```
used: none
```
) inside a day block; for week-level templates, anchor after
```
## Day 1
```
, not directly after
```
# Week N
```
.
Apply a validation loop: draft -> verify writable variables/scope/indexing -> correct -> finalize.
Preserve user intent and existing structure; change only what is required.

Basic syntax and set notation ->
```
references/language-core.md
```
Weeks, days, comments, labels, warmups, supersets, tags ->
```
references/program-structure.md
```
Built-in progression choice and arguments ->
```
references/progression-builtins.md
```
Reuse, repeat, templates, overrides ->
```
references/reuse-and-templates.md
```

progress: custom()

authoring ->

references/custom-progress-scripts.md

update: custom()

authoring ->

references/update-scripts.md

Read/write variables and indexing rules ->
```
references/runtime-reference.md
```
Liftoscript built-in functions ->
```
references/builtins-reference.md
```
End-to-end examples and fix patterns ->
```
references/recipes.md
```
Exercise name resolution ->
```
references/exercise-name-resolution.md
```
Known pitfalls and quick fixes ->
```
references/common-failure-modes.md
```

Section separators are correct and consistent for each exercise line.
Set and set-variation syntax is structurally valid (including grouped load/timer behavior when used).
Script assignments target writable variables for the selected context.
Script assignment literals use valid types for the target variable.
Timer literals follow context rules (
```
45s
```
in exercise text, numeric in scripts).
Rep tokens are numeric counts (no
```
s
```
suffix on reps; use
```
3x60
```
, not
```
3x60s
```
).
Percent load tokens in exercise-line set syntax are interpreted as
```
% of 1RM
```
.
Units are resolved from context/user preference or defaulted to
```
kg
```
.
No forbidden unit mismatches in generated output.

Exercise naming resolves against

references/exercise-list.md

(bare

Exercise Name

or explicit

Exercise Name, Equipment

When user intent specifies equipment/variant, explicit
```
Exercise Name, Equipment
```
is an exact match in
```
references/exercise-list.md
```
.
When user intent does not specify equipment, alias-only
```
Exercise Name
```
is an exact bare-name match in
```
references/exercise-list.md
```
.
If exercise is not in built-in list, output remains valid and mentions manual add path in Liftosaur.
In full mode output, template lines (
```
used: none
```
) are placed after
```
## Day 1
```
(or another explicit day), never directly under
```
# Week N
```
.