Skill Tracer
Skills › BMAD Method › bmad-toolsmith › skill-creator-loop.md

skill-creator-loop.md

23 lines · skills/bmad-toolsmith/approaches/skill-creator-loop.md at bd70c64

Open on GitHub

Most marks come from matching words in the file, so treat them as pointers.

1# Skill-creator loop
2
3Anthropic's skill-creator loop with BMad's shapes, packaging and harness. Its outcome: a skill iterated on measured results, its description tuned by a trigger eval.
4

Skill-creator loop

Anthropic's skill-creator loop with BMad's shapes, packaging and harness. Its outcome: a skill iterated on measured results, its description tuned by a trigger eval.

Raw lines 1 to 4
5## Draft
6
7Scaffold (`uv run {skill-root}/scripts/init_skill.py --name <name> --dest <parent folder> --shape <shape> --description "..."`, with `--bmod <record> --source <update_source>` for a member), load `shapes/<shape>/shape.md` and the module shape the read-back's registration line names, write `SKILL.md` to `canon.md`. Make the description deliberately specific, because widening one that stays quiet is easier than narrowing one that fires everywhere.
8

Draft

Scaffold (uv run {skill-root}/scripts/init_skill.py --name <name> --dest <parent folder> --shape <shape> --description "...", with --bmod <record> --source <update_source> for a member), load shapes/<shape>/shape.md and the module shape the read-back's registration line names, write SKILL.md to canon.md. Make the description deliberately specific, because widening one that stays quiet is easier than narrowing one that fires everywhere.

Raw lines 5 to 8
Runs code scripts/init_skill.py :7
9## Measure
11Write 2 to 3 realistic prompts into `{target}/evals/cases.json` with empty rubrics, from the user's real requests. Load `approaches/eval-loop.md` and run it to a stop condition.

Measure

Write 2 to 3 realistic prompts into {target}/evals/cases.json with empty rubrics, from the user's real requests. Load approaches/eval-loop.md and run it to a stop condition.

Raw lines 9 to 12
13## Description round
15If `{target}/evals/triggers.json` is absent, write it to the spec in `ship.md` under Trigger eval. Then invoke the `bmad-eval` skill in trigger mode with its description-optimization loop on `{target}`; it splits, measures, revises and picks. Apply the winning description.

Description round

If {target}/evals/triggers.json is absent, write it to the spec in ship.md under Trigger eval. Then invoke the bmad-eval skill in trigger mode with its description-optimization loop on {target}; it splits, measures, revises and picks. Apply the winning description.

Raw lines 13 to 16
Names a skill bmad-eval :15
17## BMad's step
19The customization decision: does a team need to swap a rule or step without forking? Then `customize.toml` and the rendered shape, else nothing. The canon pass is the lens round in `ship.md`.

BMad's step

The customization decision: does a team need to swap a rule or step without forking? Then customize.toml and the rendered shape, else nothing. The canon pass is the lens round in ship.md.

Raw lines 17 to 20
21## Then
23Load `ship.md`.

Then

Load ship.md.

Raw lines 21 to 23