Skip to content

elithril/blender-kiln / plugin

Makes and fixes 3D assets in Blender, for games, the web and AR. Use when the user wants a 3D model, prop, character or GLB/glTF/FBX/USDZ file made — from a text brief, a reference image, a free marketplace (PolyHaven, Sketchfab) or AI generation — AND when they bring an existing 3D file to fix: too heavy or too many polygons (optimize, compress, LODs), broken normals, scale or origin, materials lost on glTF export, conversion between GLB, FBX and USDZ, texturing, rigging for animation, animating it (walk and idle cycles, text-to-motion, looping clips, articulated props), or inspection. Drives Blender through a Blender MCP server (ahujasid's or the official Blender Lab one). Batch mode produces many consistent assets unattended. Kiln Studio (`/kiln studio`) opens the project's models in the browser, where the user reviews them and pins notes for Claude to act on.

FileKindSize
.claude-plugin/plugin.jsonconfig1.2 KB
.mcp.jsonconfig236 B
LICENSEother1.0 KB
NOTICEother2.5 KB
SKILL.mdskill md29.0 KB
references/ai-generation.mdmarkdown21.0 KB
references/animation.mdmarkdown21.6 KB
references/batch-mode.mdmarkdown19.5 KB
references/blender-mcp.mdmarkdown9.4 KB
references/characters.mdmarkdown52.8 KB
references/cli-tools.mdmarkdown6.5 KB
references/commands.mdmarkdown10.7 KB
references/export-targets.mdmarkdown14.3 KB
references/log-format.mdmarkdown2.2 KB
references/naming-conventions.mdmarkdown5.8 KB
references/photo-projection.mdmarkdown6.3 KB
references/pipeline-brief.mdmarkdown4.2 KB
references/pipeline-import-cleanup.mdmarkdown6.3 KB
references/pipeline-optimize-export.mdmarkdown6.4 KB
references/pipeline-rig-animate.mdmarkdown5.8 KB
references/pipeline-source.mdmarkdown9.7 KB
references/print.mdmarkdown41.5 KB
references/reference-fidelity.mdmarkdown31.8 KB
references/review-rounds.mdmarkdown10.4 KB
references/setup-install.mdmarkdown2.8 KB
references/setup.mdmarkdown9.2 KB
references/sourcing-strategy.mdmarkdown6.1 KB
references/studio.mdmarkdown38.9 KB
references/texturing-strategy.mdmarkdown14.4 KB
references/topology-rules.mdmarkdown3.9 KB
references/uv-materials.mdmarkdown6.8 KB
references/validation-checklist.mdmarkdown10.5 KB
studio/approval_hook.pyscript14.2 KB
studio/blender_live.pyscript2.3 KB
studio/checks.pyscript38.1 KB
studio/estimate.pyscript4.2 KB
studio/example/barrel/barrel_final.glbbinary111.9 KB
studio/example/barrel/barrel_log.mdmarkdown445 B
studio/handoff.pyscript17.8 KB
studio/mcp_server.pyscript32.7 KB
studio/phoneshare.pyscript17.1 KB
studio/server.pyscript47.9 KB
studio/setup_status.pyscript7.5 KB
studio/static/app.jsscript241.4 KB
studio/static/apple-touch-icon.pngbinary20.6 KB
studio/static/fonts/OFL-JetBrainsMono.txtmarkdown4.3 KB
studio/static/fonts/OFL-Sora.txtmarkdown4.3 KB
studio/static/fonts/jetbrains-mono-latin-ext.woff2binary11.3 KB
studio/static/fonts/jetbrains-mono-latin.woff2binary30.6 KB
studio/static/fonts/sora-latin-ext.woff2binary11.8 KB
studio/static/fonts/sora-latin.woff2binary24.6 KB
studio/static/i18n.jsscript6.2 KB
studio/static/i18n/fr.jsonconfig101.2 KB
studio/static/icon-32.pngbinary2.6 KB
studio/static/icon-64.pngbinary6.6 KB
studio/static/icon-96.pngbinary12.8 KB
studio/static/index.htmlother39.5 KB
studio/static/style.cssother114.5 KB
studio/static/tutorial.jsscript49.0 KB
studio/static/tutorial/barrel_after.glbbinary117.3 KB
studio/static/tutorial/barrel_before.glbbinary111.9 KB
studio/static/tutorial/fixtures.jsonconfig13.5 KB
studio/static/tutorial/lantern_1_blockout.glbbinary83.1 KB
studio/static/tutorial/lantern_2_details.glbbinary101.8 KB
studio/static/tutorial/lantern_3_materials.glbbinary102.0 KB
studio/static/tutorial/lantern_4_final.glbbinary141.1 KB
studio/static/tutorial/make_tutorial.pyscript4.9 KB
studio/static/viewer.jsscript41.4 KB
studio/studio.pyscript29.2 KB
studio/studiolib.pyscript77.8 KB
studio/test_studio.pyscript90.2 KB
studio/test_worker.pyscript31.5 KB
studio/worker.pyscript56.8 KB
studio/worker_blender.pyscript2.7 KB
tools/autorig.pyscript44.1 KB
tools/bind_rigid_parts.pyscript3.6 KB
tools/blender_mcp_start.pyscript1.5 KB
tools/check_usdz.pyscript1004 B
tools/cutout.swiftscript2.0 KB
tools/export_outputs.pyscript5.2 KB
tools/fidelity_check.pyscript39.0 KB
tools/fit_coupon.pyscript11.3 KB
tools/fixtures/skintokens_raw_child.glbbinary88.8 KB
tools/fixtures/skintokens_skeletons.jsonconfig6.5 KB
tools/flatten.pyscript14.9 KB
tools/glb_to_fbx.pyscript2.1 KB
tools/glb_to_usdz.pyscript6.2 KB
tools/motion_loop.pyscript5.9 KB
tools/motion_measure.pyscript15.2 KB
tools/motions/LICENSE-image-to-3dlab.txtmarkdown11.1 KB
tools/motions/cheer.npzbinary48.4 KB
tools/motions/cheer_b.npzbinary48.1 KB
tools/motions/dance.npzbinary78.2 KB
tools/motions/dance_b.npzbinary78.9 KB
tools/motions/death.npzbinary45.9 KB
tools/motions/death_b.npzbinary46.5 KB
tools/motions/fight_stance.npzbinary60.3 KB
tools/motions/fight_stance_b.npzbinary61.3 KB
tools/motions/flying_kick.npzbinary39.7 KB
tools/motions/flying_kick_b.npzbinary39.9 KB
tools/motions/hit_react.npzbinary31.9 KB
tools/motions/hit_react_b.npzbinary31.8 KB
tools/motions/idle.npzbinary51.8 KB
tools/motions/idle_b.npzbinary57.7 KB
tools/motions/jump.npzbinary39.8 KB
tools/motions/jump_b.npzbinary39.9 KB
tools/motions/kata.npzbinary77.4 KB
tools/motions/kata_b.npzbinary75.3 KB
tools/motions/kick.npzbinary32.4 KB
tools/motions/kick_b.npzbinary32.5 KB
tools/motions/presets.jsonconfig9.6 KB
tools/motions/punch.npzbinary32.0 KB
tools/motions/punch_b.npzbinary32.1 KB
tools/motions/punch_combo.npzbinary47.4 KB
tools/motions/punch_combo_b.npzbinary47.5 KB
tools/motions/roll.npzbinary46.7 KB
tools/motions/roll_b.npzbinary47.5 KB
tools/motions/roundhouse.npzbinary39.7 KB
tools/motions/roundhouse_b.npzbinary39.9 KB
tools/motions/run.npzbinary47.2 KB
tools/motions/run_b.npzbinary47.7 KB
tools/motions/sit.npzbinary47.2 KB
tools/motions/sit_b.npzbinary47.2 KB
tools/motions/sneak.npzbinary61.6 KB
tools/motions/sneak_b.npzbinary61.8 KB
tools/motions/strafe_left.npzbinary45.9 KB
tools/motions/strafe_left_b.npzbinary47.1 KB
tools/motions/sword_swing.npzbinary39.7 KB
tools/motions/sword_swing_b.npzbinary39.4 KB
tools/motions/walk.npzbinary62.8 KB
tools/motions/walk_b.npzbinary62.6 KB
tools/motions/walk_woman.npzbinary62.9 KB
tools/motions/wave.npzbinary46.3 KB
tools/motions/wave_b.npzbinary46.3 KB
tools/open_blender.pyscript5.0 KB
tools/photo_project.pyscript37.8 KB
tools/presets.pyscript16.3 KB
tools/print_check.pyscript47.2 KB
tools/quadruped.pyscript19.4 KB
tools/retarget.pyscript46.9 KB
tools/review_renders.pyscript13.8 KB
tools/review_sheet.pyscript11.0 KB
tools/skintokens-requirements.txtmarkdown2.4 KB
tools/skintokens.patchother9.2 KB
tools/template_profile.pyscript5.8 KB
tools/test_autorig.pyscript15.9 KB
tools/test_fidelity_check.pyscript20.7 KB
tools/test_fit_coupon.pyscript6.4 KB
tools/test_flatten.pyscript15.0 KB
tools/test_open_blender.pyscript1.3 KB
tools/test_photo_project.pyscript12.6 KB
tools/test_presets.pyscript17.6 KB
tools/test_print_check.pyscript33.1 KB
tools/test_review_renders.pyscript11.4 KB
tools/test_review_sheet.pyscript4.0 KB
tools/test_template_profile.pyscript3.6 KB
tools/test_uv_check.pyscript5.0 KB
tools/test_verify_docs.pyexecutablescript7.9 KB
tools/unimate.patchother3.5 KB
tools/unimate.pyscript17.5 KB
tools/uv_check.pyscript8.4 KB
tools/verify_blender.pyexecutablescript29.2 KB
tools/verify_docs.pyexecutablescript16.6 KB

tools/review_renders.py

View on GitHub
"""The review renders, headless: what SKILL.md rule 2 asks for when no Blender is connected.

    blender -b --factory-startup --python-exit-code 1 --python tools/review_renders.py -- \\
        MODEL [MODEL ...] --out DIR [--views q1,q3,front,back,top,under] [--mode both] \\
        [--close X,Y,Z,SIZE[,AZ,EL] ...] [--close-from check.json] [--res 1000] [--prefix NAME]

Every print session of 2026-10-13 had no Blender to talk to and wrote its own render script;
one measured a height from bound_box and read 104.8 mm for 90.0. This is that script once:

- MODEL: STL (read as mm), GLB/glTF or .blend (metres), as many as there are parts; they
  are shown where they are — give the assembled GLB for the assembled object;
- framed on the VERTICES in world space, never on bound_box (a turned object's box is larger
  than the object): orthographic, each view fitted to the vertices as that camera sees them —
  the subject spans the frame's width or height, 7 % of margin each side;
- --mode matcap (grey matcap: the geometry, what the printer sees), colour (the materials,
  the paint plan), or both — an STL has no material, so colour needs the GLB or the .blend
  (STLs render grey, which a fox's session took for a fault); a textured model shows its
  textures, a vertex-coloured one its vertex colours (its other parts their materials'
  colours), an untextured one its materials' colours;
- a rigged glTF's bone display shape (a 2 m sphere the importer adds) is left out of the frame;
- --views from front, back, left, right, top, under and the four three-quarters q1-q4
  (front -Y); the default pairs two opposite three-quarters, as rule 2 asks, with top and under;
- --close X,Y,Z,SIZE (mm) renders a close-up there, seen along the surface's normal at that spot
  (a nape seen "from outside the axis" was shot from the front, through the head); the line
  says HIDDEN when the camera still meets another surface first; AZ,EL (degrees, as the
  views: front -Y) choose the side yourself — a grazing look shows a dent's profile;
- a rigged model is shown in its rest pose, the pose its vertices are framed on;
  --close-from takes the thin spots `print_check` lists (`thickness.thin_at`, the first 4);
- each image is read back: a view where the subject covers under 1 % of the frame (an
  unframed or empty render) is an error, exit 1.

Writes {prefix}_{mode}_{view}.png and {prefix}_{mode}_close{n}.png; prints RENDER <path> for each.
"""
import json, math, os, sys

import bpy
import numpy as np
from mathutils import Vector

DIRS = {  # camera direction toward the subject: azimuth, elevation (degrees); the front is -Y
    "front": (0, 8), "back": (180, 8), "left": (-90, 8), "right": (90, 8), "top": (0, 89.9),
    "under": (0, -89.9), "q1": (-40, 25), "q2": (40, 25), "q3": (140, 25), "q4": (220, 25),
}
MIN_COVER = 0.01


def die(msg):
    print(f"review_renders: {msg}", file=sys.stderr)
    sys.exit(1)


def args():
    import argparse
    p = argparse.ArgumentParser(prog="review_renders.py")
    p.add_argument("models", nargs="+")
    p.add_argument("--out", required=True)
    p.add_argument("--views", default="q1,q3,top,under")
    p.add_argument("--mode", choices=["matcap", "colour", "both"], default="both")
    p.add_argument("--close", action="append", default=[], help="X,Y,Z,SIZE in mm")
    p.add_argument("--close-from", help="a print_check JSON: close-ups at its first 4 thin spots")
    p.add_argument("--res", type=int, default=1000)
    p.add_argument("--prefix")
    argv = sys.argv[sys.argv.index("--") + 1:] if "--" in sys.argv else []
    # "--close -25,-14,54,14" — a spot left of or in front of the axis — reads as an option to
    # argparse and stopped a session: glue the value on, as --close=-25,-14,54,14
    glued, k = [], 0
    while k < len(argv):
        if argv[k] == "--close" and k + 1 < len(argv) and argv[k + 1].startswith("-"):
            glued.append("--close=" + argv[k + 1]); k += 2
        else:
            glued.append(argv[k]); k += 1
    argv = glued
    a = p.parse_args(argv)
    bad = [v for v in a.views.split(",") if v not in DIRS]
    if bad:
        die(f"unknown view(s) {bad}; known: {', '.join(DIRS)}")
    for m in a.models:
        if not os.path.isfile(m):
            die(f"no such file: {m}")
    return a


def load(path):
    before = set(bpy.data.objects)
    ext = os.path.splitext(path)[1].lower()
    if ext == ".stl":
        bpy.ops.wm.stl_import(filepath=path, global_scale=0.001)       # mm -> m
    elif ext in (".glb", ".gltf"):
        bpy.ops.import_scene.gltf(filepath=path)
    elif ext == ".blend":
        with bpy.data.libraries.load(path) as (src, dst):
            dst.objects = list(src.objects)
        for o in dst.objects:
            if o is not None and o.type == "MESH":
                bpy.context.scene.collection.objects.link(o)
    else:
        die(f"{path}: STL, GLB/glTF or .blend")
    # A rigged glTF brings its bones' display shape along — a 2 m Icosphere in the
    # "glTF_not_exported" collection — and framing on it left a dog 2 % of the frame.
    shapes = {pb.custom_shape for o in bpy.data.objects if o.type == "ARMATURE" and o.pose
              for pb in o.pose.bones if pb.custom_shape}
    new = []
    for o in bpy.data.objects:
        if o in before or o.type != "MESH":
            continue
        if o in shapes or any(c.name.startswith("glTF_not_exported") for c in o.users_collection):
            o.hide_render = True
            continue
        new.append(o)
    if not new:
        die(f"{path}: no mesh in it")
    for o in bpy.data.objects:                  # framed on the rest vertices: show the rest pose, not a clip's frame
        if o.type == "ARMATURE":
            o.data.pose_position = "REST"
    return new


def closes(a, centre_m):
    out = []
    for c in a.close:
        try:
            v = [float(x) for x in c.split(",")]
            assert len(v) in (4, 6)
        except (ValueError, AssertionError):
            die(f"--close {c}: X,Y,Z,SIZE in mm, optionally AZ,EL in degrees")
        out.append((Vector(v[:3]) / 1000, v[3] / 1000, tuple(v[4:]) or None))
    if a.close_from:
        spots = json.load(open(a.close_from)).get("thickness", {}).get("thin_at", [])
        out += [(Vector(q["at_mm"]) / 1000, 0.012, None) for q in spots[:4]]
    return out


def base_colour(m):
    if m is None:
        return (0.8, 0.8, 0.8, 1.0)
    bsdf = next((n for n in (m.node_tree.nodes if m.node_tree else []) if n.type == "BSDF_PRINCIPLED"), None)
    return tuple(bsdf.inputs["Base Color"].default_value) if bsdf else tuple(m.diffuse_color)


def vertex_colours(o):
    """Workbench shows one colour source for every object: the painted ones their vertex colours,
    the others (eyes, a nose) their material's colour, written as vertex colours."""
    me = o.data
    if me.color_attributes:
        if me.color_attributes.active_color is None:
            me.color_attributes.active_color = me.color_attributes[0]
        return
    ca = me.color_attributes.new("review_colour", "BYTE_COLOR", "CORNER")
    cols = [base_colour(s.material) for s in o.material_slots] or [base_colour(None)]
    for poly in me.polygons:
        c = cols[min(poly.material_index, len(cols) - 1)]
        for li in poly.loop_indices:
            ca.data[li].color = c
    me.color_attributes.active_color = ca


def facing(meshes, V, N, at, s, centre):
    """A close-up looks at the surface along its normal there: seen "from outside the axis", a
    spot on the back's midline (a cat's nape) was shot from the front, through the head."""
    near = np.linalg.norm(V - np.array(at[:]), axis=1) < s / 4
    if near.sum() >= 20:                      # a dense skin: the mean normal around the spot
        n = Vector(N[near].mean(axis=0).tolist())
    else:                                     # a coarse mesh: the normal of the face nearest the spot
        best = None
        for o in meshes:
            inv = o.matrix_world.inverted()
            hit, loc, nor, _ = o.closest_point_on_mesh(inv @ at)
            if hit:
                d = (o.matrix_world @ loc - at).length
                if best is None or d < best[0]:
                    best = (d, o.matrix_world.to_3x3() @ nor)
        n = best[1] if best and best[0] < s else Vector()   # no surface within the frame: nothing to face
    if n.length < 1e-6:                       # no surface there: from outside the axis, as before
        out = Vector((at.x - centre.x, at.y - centre.y, 0))
        n = out if out.length > 1e-9 else Vector((0, -1, 0))
    n = n.normalized()
    if abs(n.z) > 0.9999:                     # straight up or down: the camera's up axis needs a tilt
        n = Vector((0, -0.002, math.copysign(1, n.z))).normalized()
    return n


BACKGROUND = 0.2235      # Workbench's own grey, kept so the images look as they always did


def cover(path):
    """The share of the frame the subject covers, read from the render's alpha — a corner taken
    as the background called a close-up filled by one flat face empty — then the image is put
    back on an opaque grey."""
    img = bpy.data.images.load(path)
    px = np.array(img.pixels[:], dtype=np.float32).reshape(-1, 4)
    a = px[:, 3:4].copy()
    px[:, :3] = px[:, :3] * a + BACKGROUND * (1 - a); px[:, 3] = 1.0
    img.pixels = px.ravel().tolist(); img.filepath_raw = path; img.file_format = "PNG"; img.save()
    bpy.data.images.remove(img)
    return float((a > 0.5).mean())


def main():
    a = args()
    bpy.ops.wm.read_factory_settings(use_empty=True)
    meshes = [o for m in a.models for o in load(m)]
    bpy.context.view_layer.update()
    V = np.array([tuple(o.matrix_world @ v.co) for o in meshes for v in o.data.vertices])
    N = np.array([tuple((o.matrix_world.to_3x3() @ v.normal).normalized()) for o in meshes for v in o.data.vertices])
    lo, hi = V.min(axis=0), V.max(axis=0)
    centre = Vector(((lo + hi) / 2).tolist())
    size = float(np.linalg.norm(hi - lo)) * 1.05                 # the diagonal: any view fits
    os.makedirs(a.out, exist_ok=True)
    prefix = a.prefix or os.path.splitext(os.path.basename(a.models[0]))[0]

    sc = bpy.context.scene
    sc.render.engine = "BLENDER_WORKBENCH"
    sc.render.resolution_x = sc.render.resolution_y = a.res
    sc.render.image_settings.file_format = "PNG"
    sc.render.image_settings.color_mode = "RGBA"
    sc.render.film_transparent = True             # the alpha says where the subject is (cover)
    sc.display.shading.show_cavity = False
    sc.display.shading.show_object_outline = False
    sc.world = sc.world or bpy.data.worlds.new("World")
    cd = bpy.data.cameras.new("ReviewCam")
    cd.type = "ORTHO"
    cd.clip_start, cd.clip_end = 1e-4, 100.0
    cam = bpy.data.objects.new("ReviewCam", cd)
    sc.collection.objects.link(cam)
    sc.camera = cam

    def towards(az, el):
        az, el = math.radians(az), math.radians(el)
        return Vector((math.sin(az) * math.cos(el), -math.cos(az) * math.cos(el), math.sin(el)))

    shots = [(v, None, None, towards(*DIRS[v])) for v in a.views.split(",")]
    for n, (at, s, view) in enumerate(closes(a, centre), 1):
        shots.append((f"close{n}", at, s, towards(*view) if view else facing(meshes, V, N, at, s, centre)))

    modes = ["matcap", "colour"] if a.mode == "both" else [a.mode]
    written = []
    for mode in modes:
        sh = sc.display.shading
        sh.light = "MATCAP" if mode == "matcap" else "STUDIO"
        # MATERIAL shows a material's flat colour only: a textured body rendered grey, read as a fault,
        # and so did a body painted with vertex colours (COLOR_0, the cat's fur)
        textured = any(n.type == "TEX_IMAGE" for o in meshes for m in o.data.materials if m and m.node_tree
                       for n in m.node_tree.nodes)
        painted = not textured and any(o.data.color_attributes for o in meshes)
        if mode != "matcap" and painted:
            for o in meshes:
                vertex_colours(o)
        sh.color_type = "SINGLE" if mode == "matcap" else ("TEXTURE" if textured else "VERTEX" if painted else "MATERIAL")
        sh.single_color = (0.8, 0.8, 0.8)
        for name, at, s, d in shots:
            rot = (-d).to_track_quat("-Z", "Y")
            if at is None:                    # a whole view: fit the vertices as this camera sees them
                R = np.array(rot.to_matrix())
                P = V @ R                     # columns: the camera's right, up, back
                plo, phi = P.min(axis=0), P.max(axis=0)
                mid = (plo + phi) / 2
                at = Vector((R @ mid).tolist())   # the middle of what it sees, back in world space
                s = float(max(phi[0] - plo[0], phi[1] - plo[1])) / 0.86
            cd.ortho_scale = s
            cam.location = at + d * max(size, s) * 4
            cam.rotation_euler = rot.to_euler()
            path = os.path.join(os.path.abspath(a.out), f"{prefix}_{mode}_{name}.png")
            sc.render.filepath = path
            bpy.ops.render.render(write_still=True)
            if not os.path.isfile(path):
                die(f"{path}: not written")
            c = cover(path)
            if c < MIN_COVER:
                die(f"{path}: the subject covers {100 * c:.2f} % of the frame — nothing framed")
            hidden = ""
            if name.startswith("close"):              # what the camera meets first, on its way to the spot
                hit, loc = sc.ray_cast(bpy.context.evaluated_depsgraph_get(), cam.location, -d)[:2]
                if hit and (loc - at).length > s / 4 and (cam.location - loc).length < (cam.location - at).length:
                    hidden = f", HIDDEN: the camera meets the surface {1000 * (loc - at).length:.0f} mm before the spot"
            print(f"RENDER {path} (subject {100 * c:.0f} % of the frame{hidden})")
            written.append(path)
    print(f"review_renders: {len(written)} image(s) in {os.path.abspath(a.out)}")


try:
    main()
except SystemExit:
    raise
except Exception as e:                                     # fail loud: --python-exit-code turns this into 1
    die(f"{type(e).__name__}: {e}")
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall