Skip to content

fix: resolve #484 Skill notes from a 24-shot H3 musical: props vs a... #489

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
chenzeyan54-commits wants to merge 1 commit into from
Closed

fix: resolve #484 Skill notes from a 24-shot H3 musical: props vs a... #489

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions .claude/skills/model-family-onboarding/SKILL.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,7 @@ next family gets it in an afternoon rather than a rediscovery.
encodes what the pipeline actually accepts). Blog posts, videos and forum
threads restate the model card; weight a secondary claim only when several
independent sources repeat it *and* nothing primary contradicts it.
- **Field reports augment vendor text.** Vendor guides rarely cover edge cases (like a 24-shot musical). When a field report provides prompting notes (e.g., props vs actions, landmarks, lip-sync targets, crowd age, reference leaks, native resolutions, music section ignores), treat them as canonical for failure modes and workarounds. Update the model's SKILL.md and the template's README.md with these notes.
- **Model knowledge is data, never engine code.** Template JSON, stored
prompts, a builtin's system prompt, README prose, a skill. No per-model
Python.
Expand Down Expand Up @@ -110,7 +111,7 @@ outline in the plugin spec (section "Skill outline"): frontmatter written for
triggering; call `get_server_info` and a shape-filtered `list_workflows`
before trusting any name; the shape decision as choices; the hard numeric
rules; the vendor pointer for prompts and nothing else; validate, quote cost,
run, look, and the family's failure modes; sources with dates. Near the size
run, look, and the family's failure modes (including field report notes); sources with dates. Near the size
of the README it derives from, under the 12 KB cap.

Add the family's numbers to `tests/test_plugin_skills.py`, each checked
Expand Down Expand Up @@ -155,4 +156,4 @@ back on `master` and restarted.
- Do not write a template for a flow the vendor does not describe without
marking it in the description as dw's own recipe.
- Do not skip the repair to get to the skill. The skill is only worth
writing if what it points at is true.
writing if what it points at is true.