recipe-pose-by-text
Pose any rig — humanoid or creature — by reading and writing numbers instead of guessing from screenshots — read in bend/side/twist, write the same terms, reach, grip, limit, diff, and only then look.
SKILL.mdplugins/apeiron/skills/recipe-pose-by-text/SKILL.md
GitHubPose by text (recipe stub)
This is a pointer, not the recipe. The workflow lives in the engine and always matches the build you are connected to, so load it from the running editor before you act:
ngine.recipes {name: "authoring/pose-by-text"}
Follow the numbered steps it returns, and read its warnings first. Do not work from this stub alone.
Use it when
- posing a character or creature for a still, a key, or a shot
- a hand must hold a prop, touch something, or reach a point
- matching a reference pose, or checking how far a pose is from a saved one
- a pose looks wrong but you cannot tell which joint is off
- fingers, a tail, a spine or any chain has to curl or aim
Ops it names
rig.describe_pose, rig.pose, rig.reach, rig.pose_goals, scene.space_switch, rig.grip, rig.grip_report, rig.auto_limits, rig.check_limits, rig.set_joint_limit, rig.socket_set, control.derive, control.set, scene.save_pose, scene.apply_pose, scene.key_pose, scene.attach_to_bone, scene.screenshot
Check each op's arguments with ngine.describe before the first call: the recipe is the
order and the discipline, the live schema is the spelling.
Start the session with editor.context. If ngine.recipes cannot be reached the editor is not
running or not connected: run /apeiron:status.