[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7F0X0TDPP03V1NECNMZV1EB":3,"skill-same:01M7F0X0TDPP03V1NECNMZV1EB":47},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":25,"commitSha":26,"skillDirectory":27,"forksCount":14,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":30,"ownerDescription":31,"updatedAt":32,"install":33,"resources":46},"01M7F0X0TDPP03V1NECNMZV1EB","auto-deskew",null,"Use when the user wants to automatically straighten a tilted image — scanned documents, phone-captured pages, photographed receipts, or slightly off-axis photos. Detects the dominant skew angle and rotates the image to vertical\u002Fhorizontal, optionally cropping the resulting transparent triangles. Uses ImageMagick `-deskew` for documents and a Hough-line fallback (via Python + OpenCV) for photographs. Triggers - \"straighten this scan\", \"this photo is tilted\", \"deskew before OCR\".","shotcowboystyle\u002Fag-skills","shotcowboystyle","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F62471?v=4","01M7EZTCFCZJ6K3MYW8BPM4X6R","Curtis Blanton","01M7EZTCG1S5M04J90A931X9BX",0,"MIT","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Fplugins\u002Fimage-ops\u002Fskills\u002Fauto-deskew\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: auto-deskew\ndescription: Use when the user wants to automatically straighten a tilted image — scanned documents, phone-captured pages, photographed receipts, or slightly off-axis photos. Detects the dominant skew angle and rotates the image to vertical\u002Fhorizontal, optionally cropping the resulting transparent triangles. Uses ImageMagick `-deskew` for documents and a Hough-line fallback (via Python + OpenCV) for photographs. Triggers - \"straighten this scan\", \"this photo is tilted\", \"deskew before OCR\".\nallowed-tools: Bash(magick *), Bash(convert *), Bash(identify *), Bash(command *), Bash(test *), Bash(getconf *), Bash(mktemp *), Bash(cat *), Bash(awk *), Bash(rm *), Bash(mkdir *), Bash(find *), Bash(xargs *), Read, Write\n---\n\n# Auto Deskew\n\nStraighten a tilted image. Two backends, picked by content type:\n\n- **document** (default) — `magick -deskew 40%`. Detects skew via the Radon-style projection profile that ImageMagick's deskew operator implements. Fast, robust on text\u002Fscans up to ±15°.\n- **photo** — Python + OpenCV Hough-line transform. Detects the dominant horizontal\u002Fvertical line angle (horizons, building edges, table corners) and rotates by its negative. Use for photographs without a clean text baseline.\n\n## When to use\n\n- Scanned documents that came in at an angle.\n- Phone snapshots of pages, receipts, whiteboards.\n- Slightly-tilted architectural \u002F horizon photos (`--mode photo`).\n- Pre-step before OCR — Tesseract accuracy degrades sharply past ±2° skew.\n\nDo **not** use this skill when:\n- The image is intentionally tilted (Dutch angle, artistic composition).\n- Skew is severe (>15°) — usually means the document was captured upside down or sideways. Detect orientation first (`exiftool -Orientation`, or rotate by 90\u002F180 multiples) before deskewing.\n- The image has no straight reference lines (clouds, abstract textures) — Hough will return noise; rotation will be near-random.\n\n## Inputs\n\n1. **Input** — file or directory. Required.\n2. **Mode** — `document` (default) | `photo` | `auto` (heuristic: if `magick identify -format \"%[colorspace]\"` is `Gray` or the image has very low saturation → `document`; else `photo`).\n3. **Recursive** — `--recursive` for directory descent.\n4. **Output** — default in-place suffix `_deskew`. `--output-dir \u003Cpath>` for sibling folder. `--overwrite` to replace.\n5. **Threshold** (mode=document) — `-deskew` percentage. Default `40%` (ImageMagick's recommended value). Higher = more aggressive detection, more false positives.\n6. **Max angle** — default `15°`. If detected skew exceeds this, refuse to rotate and flag the file (probably wrong orientation, not skew).\n7. **Crop** — default `on`. After rotation, crop to the largest inscribed rectangle so there are no transparent \u002F black triangle borders. `--no-crop` to keep the full rotated canvas.\n8. **Background** — fill colour for the post-rotation triangles when `--no-crop`. Default `white`. Use `none` for transparent (PNG\u002FWebP only).\n\n## Procedure\n\n1. Verify ImageMagick 7+. For `photo` mode, also resolve the plugin's uv venv and verify it has `opencv-python-headless` and `numpy` (managed by `image-ops-install-deps`; if missing, point at it):\n\n   ```bash\n   PLUGIN_DATA_DIR=\"${CLAUDE_USER_DATA:-${XDG_DATA_HOME:-$HOME\u002F.local\u002Fshare}\u002Fclaude-plugins}\u002Fimage-ops\"\n   VENV_PY=\"$PLUGIN_DATA_DIR\u002Fvenv\u002Fbin\u002Fpython\"\n   test -x \"$VENV_PY\" || { echo \"Plugin venv missing — run image-ops-install-deps.\"; exit 1; }\n   \"$VENV_PY\" -c 'import cv2, numpy' 2>\u002Fdev\u002Fnull \\\n     || { echo \"photo mode needs opencv-python-headless + numpy in the plugin venv — run image-ops-install-deps.\"; exit 1; }\n   ```\n\n   Invoke Python as `\"$VENV_PY\"` directly, matching the convention `image-ops-install-deps` establishes. Do not shell out to `uv run`.\n\n2. Enumerate inputs (JPEG\u002FJPG\u002FPNG\u002FTIFF\u002FTIF\u002FWebP). Skip and note RAW.\n\n3. **document** mode:\n\n   Measure first, then write — the max-angle guard in step 5 has to decide *before* the file is rotated. `-deskew` records what it detected in the `deskew:angle` artifact, so a throwaway pass to `null:` yields the angle without producing output:\n\n   ```bash\n   ANGLE=$(magick \"$IN\" -deskew \"${THRESHOLD}%\" -print '%[deskew:angle]\\n' null:)\n   ```\n\n   Apply the guard (step 5), then do the real conversion:\n\n   ```bash\n   magick \"$IN\" -background \"$BG\" -deskew \"${THRESHOLD}%\" \\\n     $( [ \"$CROP\" = on ] && echo \"+repage -fuzz 1% -trim +repage\" ) \\\n     \"$OUT\"\n   ```\n\n   `-deskew` applies the rotation. `-trim` with a small fuzz removes the resulting fill border when `CROP=on`. `$ANGLE` from the measuring pass feeds both the guard and the summary.\n\n4. **photo** mode (Python via the plugin's uv venv):\n\n   ```python\n   import cv2, numpy as np, sys\n   img = cv2.imread(sys.argv[1])\n   if img is None:\n       # Unreadable \u002F corrupt \u002F unsupported — exit non-zero so the caller skips\n       # the file rather than reading a \"0.0\" as \"no skew detected\".\n       print(\"unreadable\", file=sys.stderr)\n       sys.exit(1)\n   gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)\n   edges = cv2.Canny(gray, 50, 150, apertureSize=3)\n   lines = cv2.HoughLines(edges, 1, np.pi\u002F720, 200)\n   if lines is None:\n       print(\"0.0\"); sys.exit(0)\n   # Collect angles near horizontal (±30°) and near vertical (±30° of 90°)\n   angles = []\n   for rho, theta in lines[:200, 0]:\n       deg = np.degrees(theta) - 90  # 0 = horizontal line\n       if abs(deg) \u003C 30:\n           angles.append(deg)\n       elif abs(deg - 90) \u003C 30:\n           angles.append(deg - 90)\n       elif abs(deg + 90) \u003C 30:\n           angles.append(deg + 90)\n   if not angles:\n       print(\"0.0\"); sys.exit(0)\n   print(f\"{np.median(angles):.3f}\")\n   ```\n\n   That snippet is not shipped as a file — write it to a temp script at run time, then invoke it with the venv interpreter resolved in step 1:\n\n   ```bash\n   DESKEW_PY=\"$(mktemp \"${TMPDIR:-\u002Ftmp}\u002Fdeskew_photo.XXXXXX\")\"   # no suffix: BSD mktemp leaves it unrandomised\n   cat > \"$DESKEW_PY\" \u003C\u003C'PYEOF'\n   # ... the Python above, verbatim ...\n   PYEOF\n\n   if ! ANGLE=$(\"$VENV_PY\" \"$DESKEW_PY\" \"$IN\"); then\n     echo \"SKIPPED (unreadable): $IN\" >> \"$failed_log\"\n     rm -f \"$DESKEW_PY\"\n     continue\n   fi\n\n   magick \"$IN\" -background \"$BG\" -rotate \"$(awk -v a=\"$ANGLE\" 'BEGIN{print -1*a}')\" \\\n     $( [ \"$CROP\" = on ] && echo \"+repage -fuzz 1% -trim +repage\" ) \\\n     \"$OUT\"\n\n   rm -f \"$DESKEW_PY\"\n   ```\n\n   For a batch, write the temp script once before the loop and delete it after, rather than per file.\n\n5. **Max-angle guard**: applies to both modes, using `$ANGLE` from step 3 (document, measuring pass) or step 4 (photo). If `|angle| > MAX_ANGLE`, skip the file before writing anything and flag it:\n\n   ```bash\n   if awk -v a=\"$ANGLE\" -v m=\"$MAX_ANGLE\" 'BEGIN{ exit !(((a \u003C 0) ? -a : a) > m) }'; then\n     echo \"$IN: detected ${ANGLE}° — exceeds max-angle ${MAX_ANGLE}°, skipping (likely wrong orientation, not skew)\"\n     continue\n   fi\n   ```\n\n6. **Crop to inscribed rectangle** (when `--no-crop` is off and the simple `-trim` leaves uneven borders): use the analytic formula for the largest inscribed axis-aligned rectangle inside a rotated rectangle. For input `W×H` rotated by `θ`:\n\n   ```\n   new_w = (W·|cos θ| - H·|sin θ|) \u002F (cos²θ - sin²θ)   if W ≥ H\n   new_h = (H·|cos θ| - W·|sin θ|) \u002F (cos²θ - sin²θ)\n   ```\n\n   Crop centred. For small angles (\u003C5°) the difference vs. `-trim` is negligible — `-trim` is fine.\n\n7. For batch, parallelise with `xargs -P \"$(getconf _NPROCESSORS_ONLN 2>\u002Fdev\u002Fnull || echo 4)\"` — `nproc` is GNU coreutils and is absent on macOS\u002FBSD. Skip files ending `_deskew`.\n\n8. Per-file failures append to `$failed_log` and the batch continues. Report that list in the summary; a processed count on its own hides the skips.\n\n## Output\n\n- Deskewed images at the resolved paths.\n- Summary:\n  - Files processed \u002F skipped (RAW or non-image) \u002F failed \u002F unchanged (|angle| \u003C 0.1°) \u002F refused (|angle| > max).\n  - Per-mode count when mixed.\n  - Distribution of detected angles (min \u002F median \u002F max) — sanity check across the batch.\n\n## Notes\n\n- ImageMagick's `-deskew` is tuned for bilevel\u002Fgrayscale text. On colour photos it often returns 0° or noise — that's why `photo` mode exists.\n- For OCR pipelines: deskew before binarisation. After deskew, `magick \"$IN\" -threshold 50%` or hand off to `tesseract` directly.\n- If you want both content types handled in one run: `--mode auto`. The heuristic isn't perfect — review the angle distribution in the summary and re-run misclassified files explicitly.\n- This skill rotates only. For perspective correction (four-corner unwarping of a photographed document), use a dedicated tool like `dewarp` or OpenCV's `getPerspectiveTransform` — auto-deskew won't fix keystoning.\n","# Auto Deskew\n\nStraighten a tilted image. Two backends, picked by content type:\n\n- **document** (default) — `magick -deskew 40%`. Detects skew via the Radon-style projection profile that ImageMagick's deskew operator implements. Fast, robust on text\u002Fscans up to ±15°.\n- **photo** — Python + OpenCV Hough-line transform. Detects the dominant horizontal\u002Fvertical line angle (horizons, building edges, table corners) and rotates by its negative. Use for photographs without a clean text baseline.\n\n## When to use\n\n- Scanned documents that came in at an angle.\n- Phone snapshots of pages, receipts, whiteboards.\n- Slightly-tilted architectural \u002F horizon photos (`--mode photo`).\n- Pre-step before OCR — Tesseract accuracy degrades sharply past ±2° skew.\n\nDo **not** use this skill when:\n- The image is intentionally tilted (Dutch angle, artistic composition).\n- Skew is severe (>15°) — usually means the document was captured upside down or sideways. Detect orientation first (`exiftool -Orientation`, or rotate by 90\u002F180 multiples) before deskewing.\n- The image has no straight reference lines (clouds, abstract textures) — Hough will return noise; rotation will be near-random.\n\n## Inputs\n\n1. **Input** — file or directory. Required.\n2. **Mode** — `document` (default) | `photo` | `auto` (heuristic: if `magick identify -format \"%[colorspace]\"` is `Gray` or the image has very low saturation → `document`; else `photo`).\n3. **Recursive** — `--recursive` for directory descent.\n4. **Output** — default in-place suffix `_deskew`. `--output-dir \u003Cpath>` for sibling folder. `--overwrite` to replace.\n5. **Threshold** (mode=document) — `-deskew` percentage. Default `40%` (ImageMagick's recommended value). Higher = more aggressive detection, more false positives.\n6. **Max angle** — default `15°`. If detected skew exceeds this, refuse to rotate and flag the file (probably wrong orientation, not skew).\n7. **Crop** — default `on`. After rotation, crop to the largest inscribed rectangle so there are no transparent \u002F black triangle borders. `--no-crop` to keep the full rotated canvas.\n8. **Background** — fill colour for the post-rotation triangles when `--no-crop`. Default `white`. Use `none` for transparent (PNG\u002FWebP only).\n\n## Procedure\n\n1. Verify ImageMagick 7+. For `photo` mode, also resolve the plugin's uv venv and verify it has `opencv-python-headless` and `numpy` (managed by `image-ops-install-deps`; if missing, point at it):\n\n   ```bash\n   PLUGIN_DATA_DIR=\"${CLAUDE_USER_DATA:-${XDG_DATA_HOME:-$HOME\u002F.local\u002Fshare}\u002Fclaude-plugins}\u002Fimage-ops\"\n   VENV_PY=\"$PLUGIN_DATA_DIR\u002Fvenv\u002Fbin\u002Fpython\"\n   test -x \"$VENV_PY\" || { echo \"Plugin venv missing — run image-ops-install-deps.\"; exit 1; }\n   \"$VENV_PY\" -c 'import cv2, numpy' 2>\u002Fdev\u002Fnull \\\n     || { echo \"photo mode needs opencv-python-headless + numpy in the plugin venv — run image-ops-install-deps.\"; exit 1; }\n   ```\n\n   Invoke Python as `\"$VENV_PY\"` directly, matching the convention `image-ops-install-deps` establishes. Do not shell out to `uv run`.\n\n2. Enumerate inputs (JPEG\u002FJPG\u002FPNG\u002FTIFF\u002FTIF\u002FWebP). Skip and note RAW.\n\n3. **document** mode:\n\n   Measure first, then write — the max-angle guard in step 5 has to decide *before* the file is rotated. `-deskew` records what it detected in the `deskew:angle` artifact, so a throwaway pass to `null:` yields the angle without producing output:\n\n   ```bash\n   ANGLE=$(magick \"$IN\" -deskew \"${THRESHOLD}%\" -print '%[deskew:angle]\\n' null:)\n   ```\n\n   Apply the guard (step 5), then do the real conversion:\n\n   ```bash\n   magick \"$IN\" -background \"$BG\" -deskew \"${THRESHOLD}%\" \\\n     $( [ \"$CROP\" = on ] && echo \"+repage -fuzz 1% -trim +repage\" ) \\\n     \"$OUT\"\n   ```\n\n   `-deskew` applies the rotation. `-trim` with a small fuzz removes the resulting fill border when `CROP=on`. `$ANGLE` from the measuring pass feeds both the guard and the summary.\n\n4. **photo** mode (Python via the plugin's uv venv):\n\n   ```python\n   import cv2, numpy as np, sys\n   img = cv2.imread(sys.argv[1])\n   if img is None:\n       # Unreadable \u002F corrupt \u002F unsupported — exit non-zero so the caller skips\n       # the file rather than reading a \"0.0\" as \"no skew detected\".\n       print(\"unreadable\", file=sys.stderr)\n       sys.exit(1)\n   gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)\n   edges = cv2.Canny(gray, 50, 150, apertureSize=3)\n   lines = cv2.HoughLines(edges, 1, np.pi\u002F720, 200)\n   if lines is None:\n       print(\"0.0\"); sys.exit(0)\n   # Collect angles near horizontal (±30°) and near vertical (±30° of 90°)\n   angles = []\n   for rho, theta in lines[:200, 0]:\n       deg = np.degrees(theta) - 90  # 0 = horizontal line\n       if abs(deg) \u003C 30:\n           angles.append(deg)\n       elif abs(deg - 90) \u003C 30:\n           angles.append(deg - 90)\n       elif abs(deg + 90) \u003C 30:\n           angles.append(deg + 90)\n   if not angles:\n       print(\"0.0\"); sys.exit(0)\n   print(f\"{np.median(angles):.3f}\")\n   ```\n\n   That snippet is not shipped as a file — write it to a temp script at run time, then invoke it with the venv interpreter resolved in step 1:\n\n   ```bash\n   DESKEW_PY=\"$(mktemp \"${TMPDIR:-\u002Ftmp}\u002Fdeskew_photo.XXXXXX\")\"   # no suffix: BSD mktemp leaves it unrandomised\n   cat > \"$DESKEW_PY\" \u003C\u003C'PYEOF'\n   # ... the Python above, verbatim ...\n   PYEOF\n\n   if ! ANGLE=$(\"$VENV_PY\" \"$DESKEW_PY\" \"$IN\"); then\n     echo \"SKIPPED (unreadable): $IN\" >> \"$failed_log\"\n     rm -f \"$DESKEW_PY\"\n     continue\n   fi\n\n   magick \"$IN\" -background \"$BG\" -rotate \"$(awk -v a=\"$ANGLE\" 'BEGIN{print -1*a}')\" \\\n     $( [ \"$CROP\" = on ] && echo \"+repage -fuzz 1% -trim +repage\" ) \\\n     \"$OUT\"\n\n   rm -f \"$DESKEW_PY\"\n   ```\n\n   For a batch, write the temp script once before the loop and delete it after, rather than per file.\n\n5. **Max-angle guard**: applies to both modes, using `$ANGLE` from step 3 (document, measuring pass) or step 4 (photo). If `|angle| > MAX_ANGLE`, skip the file before writing anything and flag it:\n\n   ```bash\n   if awk -v a=\"$ANGLE\" -v m=\"$MAX_ANGLE\" 'BEGIN{ exit !(((a \u003C 0) ? -a : a) > m) }'; then\n     echo \"$IN: detected ${ANGLE}° — exceeds max-angle ${MAX_ANGLE}°, skipping (likely wrong orientation, not skew)\"\n     continue\n   fi\n   ```\n\n6. **Crop to inscribed rectangle** (when `--no-crop` is off and the simple `-trim` leaves uneven borders): use the analytic formula for the largest inscribed axis-aligned rectangle inside a rotated rectangle. For input `W×H` rotated by `θ`:\n\n   ```\n   new_w = (W·|cos θ| - H·|sin θ|) \u002F (cos²θ - sin²θ)   if W ≥ H\n   new_h = (H·|cos θ| - W·|sin θ|) \u002F (cos²θ - sin²θ)\n   ```\n\n   Crop centred. For small angles (\u003C5°) the difference vs. `-trim` is negligible — `-trim` is fine.\n\n7. For batch, parallelise with `xargs -P \"$(getconf _NPROCESSORS_ONLN 2>\u002Fdev\u002Fnull || echo 4)\"` — `nproc` is GNU coreutils and is absent on macOS\u002FBSD. Skip files ending `_deskew`.\n\n8. Per-file failures append to `$failed_log` and the batch continues. Report that list in the summary; a processed count on its own hides the skips.\n\n## Output\n\n- Deskewed images at the resolved paths.\n- Summary:\n  - Files processed \u002F skipped (RAW or non-image) \u002F failed \u002F unchanged (|angle| \u003C 0.1°) \u002F refused (|angle| > max).\n  - Per-mode count when mixed.\n  - Distribution of detected angles (min \u002F median \u002F max) — sanity check across the batch.\n\n## Notes\n\n- ImageMagick's `-deskew` is tuned for bilevel\u002Fgrayscale text. On colour photos it often returns 0° or noise — that's why `photo` mode exists.\n- For OCR pipelines: deskew before binarisation. After deskew, `magick \"$IN\" -threshold 50%` or hand off to `tesseract` directly.\n- If you want both content types handled in one run: `--mode auto`. The heuristic isn't perfect — review the angle distribution in the summary and re-run misclassified files explicitly.\n- This skill rotates only. For perspective correction (four-corner unwarping of a photographed document), use a dedicated tool like `dewarp` or OpenCV's `getPerspectiveTransform` — auto-deskew won't fix keystoning.\n",{"name":5,"description":7,"allowed-tools":24},"Bash(magick *), Bash(convert *), Bash(identify *), Bash(command *), Bash(test *), Bash(getconf *), Bash(mktemp *), Bash(cat *), Bash(awk *), Bash(rm *), Bash(mkdir *), Bash(find *), Bash(xargs *), Read, Write","main","f846d9a97082ff30d613b55a085cef425d822473","plugins\u002Fimage-ops\u002Fskills\u002Fauto-deskew","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle","https:\u002F\u002Fshotcowboystyle.github.io","Building smarter with AI, automation & modern tools. Exploring how far one person can go without a big team using today’s AI.","2026-11-10 04:58:27",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add shotcowboystyle\u002Fag-skills --skill auto-deskew",{"tool":39,"command":40},"bunx","bunx skills add shotcowboystyle\u002Fag-skills --skill auto-deskew",{"tool":42,"command":43},"pnpm","pnpm dlx skills add shotcowboystyle\u002Fag-skills --skill auto-deskew","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Fplugins\u002Fimage-ops\u002Fskills\u002Fauto-deskew\u002FSKILL.md",true,[],[48,54,61,68,75,82],{"skillNo":49,"name":50,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":51,"categories":52,"keywords":53,"publishedAt":19},"01M7F1YRY5S5QRJ1V0BJS130QX","no-description","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fno-description\u002FSKILL.md",[],[],{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":19},"01M7F1Y9BV2FJNQJ3NB32JS5CM","good-skill","Fixture skill that satisfies every rule.","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fgood-skill\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":19},"01M7F1XXBQ3NQRXZSS0KAGAR9V","broken-link","Links to a reference file that does not exist.","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fbroken-link\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"publishedAt":19},"01M7F1XBX4R46NR9W0HXQ7XCDX","Bad_Name","Name violates the kebab-case rule and does not match the directory.","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fbad-name\u002FSKILL.md",[],[],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7F1WNAF8WNEQ7S5HKT5VK0D","skill-authoring","Write, review, and validate portable agent skills in SKILL.md format. Use when creating a new skill, editing an existing one, or asked whether a skill follows the Agent Skills spec.","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Fskills\u002Fskill-authoring\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M7F1W30TC0XFNSF7YBRGETK2","weigh-bags","Record scale readings for each bag — empty (tare), outbound, current, final — and reconcile them against the itemised inventory to find unaccounted weight. Use when the user gives a bag weight, says how heavy something is, weighs their luggage, or asks how close to the limit they are. Appends to bags.yaml and runs scripts\u002Freconcile.py.","https:\u002F\u002Fgithub.com\u002Fshotcowboystyle\u002Fag-skills\u002Fblob\u002Fmain\u002Fplugins\u002Ftravel-packing-assistant\u002Fskills\u002Fweigh-bags\u002FSKILL.md",[],[]]